Stories as findings in collaborative research: making meaning through fictional writing with disadvantaged young people

Abstract

Working in a participatory research project with young people who are disabled, care-experienced or otherwise disadvantaged, collaborative fiction writing was a core method of hearing and amplifying their voices. We discuss how meanings were made in this iterative process of capturing resonances in the different stages of the research, resulting in the creation of stories filtered through many different participants. Through individual and joint reflections on the complex processes of constructing the 48 short stories, we demonstrate how collective storytelling can address criticisms of fictional research outputs as (in)valid social science, and argue instead that the resulting stories can be considered rigorous and faithful research findings. We suggest that these research outputs preserve and proliferate the meanings of marginalised young people, and challenge the absence or distortion of existing narratives about their lives as experienced by themselves.

Introduction

Making meaning is the aim of any qualitative research project. In participatory research, the aim is specifically to democratise the process of meaning-making by including and valuing the perspectives of people who tend to be excluded from knowledge-production (Fals-Borda and Rahman, 1991). Increasing focus on interdisciplinary and transdisciplinary research provides potential for arts and social science work to contribute to authentic and democratic knowledge creation through innovative means. This article explores how this was achieved through collaborative meaning-making in stories with children and young people in a large UK Research Council-funded participatory research project. In this study, children and young people in contact with specialist children and youth services were involved as young researchers (YRs) and as interviewees. Based on the notion that stories are central to our means of communicating with ourselves and one another (Bruner, 2002; Siegel, 2015), the project’s aim was to enhance understanding of disadvantaged young people’s perspectives through accessible fictionalised stories. Whilst arts-based representations of research findings can create more open spaces in which dialogue about meanings can occur, artists and academics have also described concerns about arts-based approaches adding additional layers of interpretation, undermining rigour and faithful representation (Boydell et al., 2016). This is particularly the case when individual interpretation is prominent. The YRs involved in the Stories2Connect (details provided in Funding section) study, however, have argued that conveying findings as stories provides them with opportunities to put their emotions into the stories told and characters described whilst protecting them from over-exposure in publicly reporting findings (Dan et al., 2019). In this article, therefore, we explore how the drafting, writing, editing, illustration and digitalisation of fictionalised stories based on qualitative research data constitute rigorous mechanisms for layering meaning. This has significance in demonstrating how collaborative creation of stories as research outputs, through implicit coding and theming comparable to an iterative process of collaborative data analysis, can preserve and proliferate the meanings of marginalised young people.

We begin with a review of the uses of storytelling in research demonstrating its relevance as a methodology with which children and young people can construct and tell meanings. Collaboration and fictionalisation, two distinctive features of our employment of stories within the operationalisation of this present study, are examined along with the innovative elements of our methodology. Our own experiences of story-making in relation to existing approaches and tensions in arts-based qualitative research are discussed.

A review of stories as research

This section gives a rationale for the use of narrative and storytelling approaches in our work, while also noting the lack of research involving analysis by and with children. We address the concern that stories are not necessarily regarded as ‘scientific’ means of conveying meaning and claim that our collaborative analysis goes some way to actively acknowledging that stories are both ‘means of knowing and a method of telling’ (Richardson, 1997: 58).

Stories as representations of life experience are not a new idea: (auto)biographies are popular methods of conveying apparent ‘truths’ in the form of a story. In research terms, ethnographies are ways of telling stories of different ways of living (Geertz, 1973; Mauss, 1947), sometimes combining fiction and nonfiction for deliberate effect. Narrative approaches to qualitative fieldwork provide means of communication which can enable interviewees and co-researchers to recount their stories (Boje, 2002; Clandinin and Huber, 2002; Coste, 1989; Frank, 2010; Goodley and Clough, 2004), and can convey information to a wide audience, including people who may not access more formal means of communication (Hinyard and Kreuter, 2007; Wahler et al., 2009). Narrative approaches in social work and community work have long been seen as potentially enabling service-users and oppressed communities to name experiences, negotiate critical moments in their lives and engage with and empower communities and vulnerable groups around issues of social justice and inequality (Christensen, 2012; Lenette et al., 2015; Martin, 1998; Roets et al., 2007; Schiettecat et al., 2018; White, 2003).

This article focuses on the uses of narrative to convey findings; little sociological work does this, and yet creative writing has much to offer as an effective means of communication (Gordon, 2008). Indeed, stories are strategically used in political and media arenas to influence policy and public attitudes, because they present accessible and emotionally engaging claims to evidence. Such stories, however, often reinforce and actively construct negative portrayals of young people (e.g. as ‘feral youths’ and ‘knife-wielding yobs’ [Gillespie, 2018]), disabled people or those who access services. Tyler (2015) discusses the subsequent effects on public opinion of TV programmes such as Benefits Street, which result in demonisation of, for example, ‘the chav’ and ‘the benefits cheat’. The cultural production of stigma is reflected in children’s literature when disabled characters are often omitted, marginalised or stereotyped (Booktrust, 2009; Dahl, 1993). But, when collaborative research celebrates young people’s ‘resistance and their criticality’, the stories generated can tell narratives of personal encounters and ‘wider sociohistorical, political and cultural events’ that can enable inclusion and challenge misconceptions (Goodley and Clough, 2004: 349). Arizpe and Styles (2010) identify the telling of stories by children and young adults as a neglected dimension of children’s literature scholarship. Our collaborative research attempts to provide counter-narratives from the inside: to tell new stories that convey data with the integrity of robust sociological analysis but which enable imaginative and emotional engagement.

Nind (2011) suggests that narrative and life-story work allows valuing participants as expert witnesses in the active telling and retelling of their own experiences, while Keats (2009) describes the benefits of multiple texts analysis in narrative research: ‘Including a variety of participant-constructed narratives . . . reflects the complexity of life experiences’ (p. 182). Yet Keats’ work does not involve the (adult) participants in the analysis, and Nind (2011) notes that more research is needed on how learning-disabled children, in particular, can engage in data analysis. In a review of research with children in contact with child welfare services, of the 78 studies included, only four described how children were included in data analysis (Kiili et al., 2019). Subsequent research (Larkins et al., 2013; Gillet-Swan and Sargeant, 2018) has shown that narrative summaries of interviews and collaborative coding can enable children to add cultural and contextual insight in the analysis of data generated with other children and young people. We take this approach two steps further, by including multiple and transdisciplinary participants – both adults and children – in the analysis, and by introducing fictionalisation. We know of no participatory research with children and young people which has used fiction to assemble and layer meanings.

Fictionalisation, representation and reflexivity

Presenting research findings as a matter of storytelling is at once an established tradition and a relatively new and challenging idea. Abbott (2007: 96) concludes that there is a place for ‘lyrical writing’ even within social science’s requirement for ‘rigor and investigative detachment’, and Smart (2010) points out the difference between evoking ‘atmospheres’ of emotion and manipulating the reader ‘towards a particular pre-determined goal’ (p. 10). Gordon (2008 [1997]: 22) argues that there is ‘more to learn about how to conjure [up social life] in an evocative and compelling way’ and advocates researchers making common cause with the objects and subjects of their research, to ‘reckon with how we are in these stories, how they change with us, with our own ghosts’ and to make visible the marginal, forgotten and repressed. Similar arguments circulate around other forms of representation of data such as the use of poetry (Carter et al., 2018).

Bridging the gap between sociology and literary fiction, novelist Toni Morrison in The Origin of Others (2017) discusses the process of transforming fact into fiction, including crafting a hopeful end to her novel Beloved, unlike the true story which inspired it. In a chapter called ‘Narrating the Other’, she concludes ‘Narrative fiction provides a controlled wilderness, an opportunity to be and to become the Other. The stranger. With sympathy, clarity, and the risk of self-examination’ (p. 91). This ‘self-examination’ is the element of reflexivity that we, as academics, are expected to bring to qualitative research, although Morrison’s reference to ‘risk’ also reminds us of the difficulty of fully acknowledging our own positions, assumptions and potential abuses of power, and the risks we might be exposing the young people to through the act of writing. Clough’s (2002) foundational work containing fictional stories from educational settings sets out to ‘lay bare’ how ‘meaning is created and communicated in research processes’, and does not shy away from acknowledging the contribution made by ‘our own selves, the ultimate sources of data’ (p. 4–5). Richardson (1994) writes, ‘Self-reflexivity unmasks complex political/ideological agendas hidden in our writing . . . desires to speak “for” others are suspect’ (p. 523). Tensions abound in the relationship between qualitative research and fiction.

Patricia Leavy (2015, 2016), discussing this relationship, is clear that there is a place for fiction-writing in the pursuit of promoting empathy through understanding and disrupting stereotypes, although Watson (2009) problematises empathy as a legitimate pursuit of research presentation. However, Leavy’s focus is largely on how to construct fiction which does this effectively, rather than exploring the methodological tensions and ethical responsibilities for all participants in using fiction in qualitative research. Creating a list of ‘traditional qualitative evaluative criteria transformed for fiction as research’, Leavy (2016: 79) replaces ‘validity’ with ‘it could have happened’ and ‘rigor’ with ‘aesthetics’; while ‘trustworthiness’ becomes ‘resonance’, ‘authenticity’ becomes ‘verisimilitude’ and ‘reflexivity’ is transformed into ‘author’s personal signature’. It is significant in the context of our own work that the writers to whom Leavy refers as exemplars of the art are individual authors who are also academics. So, to what extent are these evaluative criteria relevant for a participatory project like ours where children are co-authors of stories? In an academic article submitted for publication (see Satchwell and Davidge, 2018), one of the reviewers chose to assess the embedded co-constructed short story using Leavy’s (2016) criteria; the story ‘failed’ on several counts. The inclusion of precise details from our research data at the expense of plot devices and sensory imagery in our fictional story had resulted, according to the reviewer, in a story of diminished artistry. Such dilemmas are addressed further in later sections of this article.

While Clough’s (2002) work starts to bring together ethnographic and literary traditions (p. 12), ours goes a little further in exploring co-construction and the attempts to keep the voices of the original participants – attempting to speak not ‘for’ others but with them. Our research recognises that ‘the realm of meaning is best captured through the qualitative nuances of its expression in ordinary language’ (Polkinghorne, 1988: 10), while also accepting that the hermeneutic reasoning required by linguistic data ‘does not produce certain and necessary conclusions’ (Polkinghorne, 1988: 7). However, we suggest that the collaborative, layered, iterative hermeneutic processes involved in our research analysis strengthen – rather than weaken – its claim to knowledge.

Methodology

All participatory research with children and young people involves adults as well; we use the term collaborative research as this study was co-initiated by young people and adults, rather than an adult-initiated study in which children participate. The nature of this collaboration is best analysed as a lattice (Larkins et al., 2014) which recognises that different members of the team and participants exerted different levels of influence at different stages of the project (e.g. initiating, generating data, analysing findings, creating outputs, pursuing dissemination and influencing change in life-worlds). It is impossible to identify exactly how decisions were made as power in interpersonal and intergenerational research settings happens through a wide range of verbal and non-verbal utterances. However, influence occurred through different forms of dialogue and story-telling as discussed in our findings. Rather than the division being solely generational, influence appeared to vary between different members of the team according to interests, time available for the project and their previous experience.

The Stories2Connect project began with an existing partnership between one post-1992 university, a UK children’s charity, and an ongoing young researcher group hosted by the university, in which young people identify issues to research in collaboration with academics. When the project began, young people involved with the young researcher group and charity collaborated with the academics to select and recruit an adult research fellow. The core group of YRs was then expanded through invitation to young people involved in participation groups for children and young people in contact with specialist children’s services. This led to the establishment of a group of 13 YRs (aged 13 to 24 years). These young people (5 female, 8 male; all White European) all had experience of being in alternative care (6), being disabled (12 had autism, Down Syndrome or learning difficulties) and being young carers (2). The adult researchers (ARs) comprised seven academics from education, health, literacy, literature, social work, psychology and design, most but not all with experience of working in participatory ways with young people. Each of us viewed the young participants and the potential data produced by and with them from slightly different personal and disciplinary perspectives; for example, colleagues in Literature or Design had somewhat different priorities from those in Social Work or Health, as we highlight later in the article. The team learnt from one another through a constant to-ing and fro-ing between different positions, but (arguably) the adults learnt most from the young people.

Over 3 years, the ARs and YRs team had regular monthly evening meetings, with additional daytime meetings and workshops at weekends and during school/college holidays. Despite the turbulent backgrounds of some of these individuals, only two left the group before the end of the 3 years, due to changes in home circumstances, and one of these is still in touch with us.

ARs and YRs experimented with methods for conducting interviews, considering issues of ethics, confidentiality and anonymity and through this shared experience devised and agreed on a set of questions. The YRs interviewed one another and then moved out into the communities they identified as their own to interview other young people who faced physical, social, economic, educational or emotional challenges in their lives. In total, the team conducted 95 interviews with 65 YRs and participants, using the following four main types of interviews:

1. YRs interviewing one another (18)

2. YRs interviewing other young people with an adult researcher present (29)

3. ARs interviewing young people with a YR present (6)

4. ARs interviewing young people without a YR present (9).

In addition, the ARs conducted 33 evaluation interviews with the YRs about their experience of participation in the present study and other projects. Overall the interviews took place in a range of settings: colleges, schools, youth groups, homes, cafés, supported housing, university rooms or social spaces.

All interviews were transcribed, and most (but not all) became sources for stories. Detailed records were also kept of stories that were made up using more deliberately creative fictional methods with young people in project workshops, or stories that were constructed ‘on the hoof’ in locations like a station platform or a fast-food restaurant. Through conventional and creative analysis, these interviews then became the source of data for storymaking.

The fictionalisation of the stories became an intricate and revealing process, which highlighted the meanings that could be made through interaction between qualitative data and creative interpretation. Additional stages of the project, including the processes of illustrating and animating the stories, added further layers of interpretation and intrigue, as did the creation of digital storytelling machines to disseminate the stories. Interpretation by readers and viewers are further considerations explored elsewhere (e.g. Satchwell, 2019). Our attention in this article though is on the processes of storymaking. Therefore, in exploring meaning-making in the stories, we focus on influence within this fictionalisation process.

To reflect on this process of analysing data and story-telling, the four academics who took active roles in storymaking (the authors of this paper) wrote vignettes to share our personal experiences and reflections; we also collected reflections from writers external to the project team. One limitation of this article is the absence of YRs’ own perspectives on the extent to which they felt the process enabled democratisation of meaning-making through collaborative data analysis and storytelling. However, articles they have contributed to and co-authored demonstrate that they feel they had some control over the storytelling and they felt represented in the stories (Satchwell and Davidge, 2018; Satchwell, 2018; Dan et al., 2019). In this article, we reflect on how we facilitated these processes, with attention to our own influence. We present an overview of the processes of generating data and conventional analysis, followed by a synthesis of our reflections on the storymaking component, discussed in the light of relevant literature.

Generating data and conventional analysis

Our original research design included peer interviewing using an interview schedule designed by young people as our main method of collecting data but, in practice, the research evolved and grew according to the human interactions between and among adult and child participants. The YRs, particularly two of the more vocal ones, argued that we should focus the interviews on highlights, as these were things that children and young people would want to talk about. The academics with backgrounds in literature and the children’s authors who advised us wanted details of characters, settings and a crisis to be resolved, whereas the desire to safeguard both the interviewees and interviewers, most of whom had experienced significant traumatic events in their lives, meant that some academics were influenced by a protectionist and strengths-based approach. Different ways of collecting data therefore combined to reflect these priorities.

Individual YRs and ARs made decisions in the field about which elements of the interview guide to use (highlights, challenges, characters and personal interests) and individual interviewees made decisions about how to respond. More direct and structured interviews, on the other hand, meant the (young or adult) researchers who were less confident were able to ask specific questions and create opportunities for participants to articulate meanings. Distance and space without questions also allowed for participants’ meanings to emerge: ‘Bit by bit I gleaned pieces of her life-narrative that she offered up while we were engaged together in artwork or some other activity, her story coming sideways rather than face on in an interview situation’ (an AR). Occasionally ARs or YRs also worked with young participants to create a fictional story, giving a freedom to ‘invent’ a crisis or an antagonist, inevitably drawing on the young people’s own lives, but at a safe distance – a form of ‘externalising the problem’ (White and Epston, 1990). As a result, the data we collected varied considerably and when we moved to craft the fictional stories, our starting points included interview transcripts or the outline of a young person’s imaginative story.

Three ARs analysed the transcripts of interviews using NVivo (QSR 2016) qualitative analysis software, using themes that had been identified or agreed by the YRs (e.g. aspirations, challenges, people who help and what young people do) and additional themes suggested by the ARs based on previous participatory studies with children (personal resources, social resources, emotional support and family life). Through reading and rereading the transcripts, three ARs identified numerous subsets of these themes, related to patterns of commonality and difference between the transcripts (Fraser, 2004). In the same way that our ideas differed about suitable questions to ask, what was sought in the analysis was informed by our starting points as researchers: our disciplines, previous work and academic interests, as well as following ‘hunches’ and ‘ghosts’. Such influences were discussed in group meetings between the three ARs involved in analysis, but inevitably were not fully acknowledged all of the time.

ARs then fed back the framework of these themes and subsets to young people in four ways. First, we conducted a collaborative, inclusive method of analysis, wherein YRs sorted snippets of interviews according to what they felt was important. They reviewed interview transcripts and identified significant elements using the question ‘What strikes me?’ Arguably the framework was still determined to some extent by themes identified at the outset. For example, the initial research aim was to provide stories of ‘resilience and transformation’, and therefore the categories reflect an examination of how young people overcame adversities in their lives. However, the YRs co-created interview schedules, and certainly their contributions in the interviewing and co-analysis stages triggered new categories, such as ‘bullying’.

Using their selection of themes, YRs created story-dice, which they threw to create storylines enacted through drama at a workshop. Further, the ARs and YRs played a game that presented all the themes and summaries of content from the emerging analysis with another group of young participants. This led to concepts being questioned and new subthemes of ‘role models’ and ‘feeling safe’ being added to the NVivo analysis process. Revised themes were then explored with the YRs playing the same game, leading to themes in NVivo such as ‘participation’ and ‘social action’ being replaced with ‘making a difference’ through being heard and other actions.

In summary, young people influenced some of the conventional data analysis resulting, for example, in changes in the concepts and wording and generating clearer themes. The data within NVivo were then recoded. The themes and subthemes ‘emerged’ from the data in a semi-grounded theory approach within a thematic structure through a process of deductive and inductive coding involving some of the ARs and some of the YRs and participants. The adults sought to set aside different themes which might have ‘emerged’ had we chosen to theme the data according to our own different priorities. However, the original questions we asked have irretrievably influenced the ‘findings’. Our assumption that young people had achieved something and had overcome challenges and barriers predetermined what would be presented as data and become available for analysis. As in all qualitative data analysis, the analysis is in part driven by what we hope and expect to find: ‘Every enquiry is guided beforehand by what is sought’ (Heidegger, 1962: 24). Our ghosts do not subside whatever our means of analysis.

Constructing the stories: whose meaning is it?

Alongside interviewing and conventional data analysis, story-making began through a variety of different means. To achieve this we worked with the YRs and we engaged 25 writers and 10 illustrators from a society in the local area who specialised in writing fiction aimed at children and young people. They were interested in being involved in a ‘worthwhile’ project, and the challenge of finding authentic ways of fictionalising young people’s own accounts of their lives. Eight other writers and illustrators were students from three universities. The authors of this paper also had roles as writers, reviewers and editors, writing several stories and painstakingly reviewing multiple drafts of stories with reference to both aesthetics and authenticity.

In summary, the four story-telling methods identified were:

1. ARs working with individual young people to create stories (e.g. using a story arc).

2. Young people creating stories together in groups, facilitated by an adult researcher (e.g. using a story-making game; story-making dice; story bags).

3. Creative writers who were also researchers constructing stories from different sources, including fieldnotes and transcripts of interviews.

4. Creative writers who were external to the project using ‘story ingredients’ provided by researchers.

The ARs selected transcripts or a combination of excerpts and additional relevant information to send to the writers. Writers were therefore guided by the selections made by ARs (informed by the NVivo analysis and YRs’ striking content), but also with an indication that there was space to invent ‘missing’ content about the young people themselves where young participants or YRs had not chosen to share or explore personal details.

One tension in this meaning-making process concerned relevance. One of us described creating a story with a young man with autism and feeling discomfited by the process: ‘Prompts that seemed important to me such as “tell me more” or “why did that happen” were often dismissed as being “irrelevant” or responded to irritably as if I should be able to keep up. I felt inept and out of my depth’. Thereafter she ‘set aside her expectations’, raising the notion of dissonance in assumptions between adults and young people with autism and other disabilities, and highlighting that we needed to play by different (their) rules. If we had not embraced their differences there are several stories that would have been presented more conventionally. Leavy’s (2016) evaluation criteria are based on values espoused by a section of readers, not all. Some stories were therefore written which may seem difficult to link to the project aims, but which were of clear relevance to the YRs.

A second tension, between authenticity, ethics and aesthetics, is illustrated in the following three examples, each resulting in a more faithful representation of young people’s meanings, if not their words:

1. Sending a draft story to a colleague drew the comment that the meaning conveyed by one line was unrealistic. This turned out to be a direct quote from a young person, so we left it.

2. A story that was deemed to be too much like a monologue, using almost entirely direct quotes, was rewritten. On reflection, the author felt that there was no significant difference in meaning, and perhaps the change in style conveyed more of the sense of shock and anxiety that was silent in the transcript (Spyrou, 2016).

3. A writer’s main aim was to present an interviewee’s experiences ‘as faithfully as possible’ saying: ‘I tried to imagine I was in conversation with her and balancing her wish for representation with my concerns about anonymity and audience engagement . . . It was sent for literary review and this highlighted the need to root the story more in place. Interestingly I’d removed place from the story in order to anonymise!’

These examples illustrate how we attempted to resolve the tensions between the different criteria of qualitative research and fiction as they emerged in the process of writing.

The tension in understanding what is realistic was also felt when the lived experiences conveyed in stories were outside standard representations of children and young people’s lives. One of our interviewees stated that he loved housework and cocktails. Another experienced a train of deaths of multiple relatives and support workers. These details may seem unrealistic but to omit them is a distortion; some children’s conditions lead them to somewhat unusual obsessions, and the challenges within families and social disadvantage caused by inadequate social care mean that other children experience intense personal crises without consistent support. While the editing process reduced the death count by one in an attempt to lighten the gloom, we therefore wrote a story that features multiple deaths, countered by the growth of an oak tree. In another story, housework and cocktails appear alongside an account of how the courage to vote in an election for the first time arises for a character called Lenny, after he meets a girl in a nightclub:

First thing she says to him is: ‘Love your bling, man - you’re, like, shining - you’re like a star.’ The word ‘Love’ makes him feel warm and weak, and he says to the side of her face: ‘I like housework.’ She laughs loudly with her mouth wide open. Lenny thinks, when she laughs, she is even prettier.

• ‘Housework? Right? Cool!’ she says.

• And then she says: ‘Dance, yeah?’

• Lenny takes a deep breath and makes eye contact with Anya. He thinks about the two cocktails he’s bought for himself because he didn’t want to queue up twice: one for now, one for later. They aren’t the best cocktails ever. In fact, Lenny knows he makes much better - but they cost shed loads.

We suggest that inclusion of unusual narratives and details is exactly what makes these stories realistic.

In order to represent young people’s meanings, the adult writers referred to ‘sensing’ the young people in different ways:

‘I felt that XXX was beside me, editing my more florid prose, paring things back, and helping me to fill the gaps’;

‘I would approach a particular story by immersing myself in all the interviews with or concerning one or more young person, along with my own experiences and observations of those individuals, until I felt I “inhabited” the characters enough to write ‘truthfully’ about them’;

‘This story, concerning a girl who cares for her siblings in the face of her mother’s depression, sat in my mind for months before I started writing’.

These examples of editing oneself back, immersion and sitting with ideas over time all imply attempts at a visceral connection with the young people in the process of creating authentic characters.

We contributed our own meanings, as ghost-writers, evoking the spirit of the young people, while also seeing ghosts of our own childhoods, people we have known and our professional experiences (Gordon, 2008). But we also feel ghosts of our selves when reporting qualitative interviews in a more conventional sociological style; any academic endeavour is an ongoing dialogue with personal and professional experience. Connections with ghosts were emotional: a co-created story with a girl in foster care who had experienced serious assault in her earlier life led one writer to seek support herself ‘because I was so moved by her story’. Other writers refer to being inspired by depictions of relationships, events or images which had connections for them: ‘picking up on details’, ‘several bits jumped out at me’ and ‘I began to get into the mind of a young person’. These writers commented on the different but ultimately rewarding way of working: ‘the transcript was so insightful – it was a great way to drive forward a story I would not have ordinarily written’; ‘At first I found it difficult to find the right balance between fictionalising the information and keeping the story true to its original inspiration’. They also referred specifically to influences from their own lives, for example ‘the protagonist’s friend is based on my son’; ‘having had an alcoholic parent helped me to relate to the mother in the story’. These examples of resonance between the young people and their lives, via the transcripts, and the writers, are examples of stories connecting people. Our attendance to these personal emotional aspects was inescapable and as Harrison et al. (2001) remind us, this attendance is a ‘criterion of trustworthiness’ in feminist research (p. 326).

Opportunities for readers’ own meaning-making were created by participants and writers, leaving gaps and distancing. For example, at least two stories contain obscure reference to experience of abuse, which may be undetectable to most readers. The ambiguity or ‘gap’ was deliberate so that readers can make their own meaning of the source of fear or anxiety. Humour was used as a means of distancing both young participants and future readers from traumatic events. For example, a hilarious story-making activity at a youth club involved somewhat bizarre stories emerging during a game involving coloured paper and beanbags, choosing characters, problems, support mechanisms and outcomes. Writing up the stories created by the young people, the academic researchers retained topics of drug-dealing, unemployment and potential violence, while also including humour. The construction of another story, which details the mounting fear of a vulnerable girl being approached by a gang of lads, is resolved with a comical final scene: a deliberate attempt to draw in a reader but also a device for keeping any future (child) reader safe.

One recurrent ambiguity arose from writing stories from the perspectives of the young people themselves: they do not dwell on a ‘diagnosis’, ‘condition’ or a ‘label’, and Satchwell and Davidge (2018) argue for the importance of focusing on human characteristics rather than medical or psychological diagnoses. Writers responded through storytelling techniques: there is no reference in Lenny to SEND, and yet as readers, we recognise his idiosyncrasies, anxieties and triumphs as a human being. Another writer wrote in her reflection: ‘I haven’t actually mentioned the disability of the main character – I don’t know what that is specifically, but I’m not sure it needs more detail on that front’. A writer whose subject was a young carer said, ‘The main challenge was what was missing – the interviewee didn’t want to talk about his problems’. The resulting story was entitled ‘Just getting on with it’: a refrain that echoes this silence. A writer presented with data referring to anxieties about independent travelling commented: ‘Any normal journey (e.g. bus or train) didn’t strike me as an exciting or engaging thing to write about’, but when starting to ‘write it as sci-fi’, ‘the story instantly flowed’.

Correspondence between the diversity of the data generated and the fictionalised stories was monitored by assessing them against the conventional coding, checking to ensure we were covering all themes, making strategic stories where necessary and ensuring that stories picked up on the outlying themes as well as the more common experiences. For example, at one point we realised we needed to redress the omission of a story about children being groomed by drug-dealers and a story that acknowledged a positive outcome for a family experiencing foster care: these issues had arisen in interviews but had not yet been incorporated into stories. At the same time, we were monitoring aesthetic ‘quality’ of the stories by reading, drafting and re-writing, testing them out on young people, continuing to hone wording and structure. We acknowledge that many more fictional stories are needed to do justice to the narratives gathered.

Discussion: co-created fiction as an analysis of qualitative data leading to findings

The presentation here of our means of producing the stories is itself an acknowledgement of the limitations of this kind of research analysis. By openly discussing the choices, influences and messiness of our co-construction processes, we are inviting accusations of a lack of scientific rigour. However, our point is that all analysis, however conventionally ‘rigorous’, is subject to human influence. Data are always open to reinterpretation and any finding also involves a covering up. Rather than pretend otherwise, we agree with Stronach and MacLure (1997) that every opening relies upon a closing and vice versa. Therefore, we make no claims that we are presenting the findings through these stories; rather we suggest that co-creating fiction is a way of co-constructing and co-conveying meanings in collaborative qualitative research.

Leavy (2016) suggests that in fiction-based research, the term ‘findings’ is ‘simply irrelevant’ (p. 78) and proposes alternative means of assessing and evaluating the research. She suggests that the fiction should be assessed according to the goals of the research: for example, in our research, we aimed to challenge stereotypes by presenting authentic stories of young people’s lives. The effectiveness of the stories in achieving these aims is something to be considered once the fiction is produced and this will be assessed over the long term through future research. But we know already that YRs, parents and conference participants have said that these stories are experienced by some people as sensitive representations of their lives or the lives of their children which they feel could change attitudes.

Leavy’s evaluation criteria include ‘It could have happened’. But who is to judge this? We suggest that this evaluation is a distraction when research seeks to bring to the fore knowledge and meanings that have been obscured by dominant and discriminatory public representations of children and young people. An alternative criterion is trustworthiness. We have shown that all the writers drew on their own experiences, feelings and values in the construction of the stories, while also vigorously attempting to retain those of the young people: ‘There’s inevitably a little bit of me in there but hopefully not too much’. Trustworthiness is a characteristic of good qualitative research (e.g. Harrison et al., 2001; Brantlinger et al., 2005), but also of a good researcher (Finlay 2002; Mauthner and Doucet, 2003). If we are drawing parallels here between researchers and writers in the process of constructing stories, the same criteria apply. Harrison et al. (2001) drawing on Lather (1991), relate trustworthiness to reciprocity. The to-ing and fro-ing of our transcripts and drafts is a form of reciprocity, where the stories are ‘multiples, not monographs, but clusters of many texts’ (Nespor et al., 1995: 61), increasing the possibility of us being able ‘to engage politically with all of our relevant audiences’ (Nespor et al., 1995: 61).

The trustworthiness of fictionalised story-telling must therefore be equated with meaning-making in the process and products of research. Techniques to enable this include: YRs helping to develop and implement interview schedules; involvement of YRs and participants in creative data analysis; recruiting writers committed to authentically representing young people; inclusion of themes that seem irrelevant; line by line reflection on the authenticity of meanings conveyed alongside consideration of whether this needs to be in participants’ own words; inclusion of details that seem unlikely and unrealistic; naming and valuing the ghost meanings conveyed by writers; allowing gaps for readers’ meaning-making; writing from the perspective of young people (rather than their label); and monitoring the correspondence between stories and analysed data.

To return to Leavy’s criteria, ‘resonance’ figures highly as equating with both ‘validity’ and ‘trustworthiness’ (2015: 79). For us, it was of utmost importance that ‘resonance’ occurred with the YRs and their interviewees themselves, before even considering resonance with a future reader. These young people signalled resonance by laughter, nodding, smiling and sometimes complete transfixed silence as a story was read out (an unusual phenomenon in this group). They registered dissonance by suggesting changes, frowning, arguing or wandering away during the telling. These signals – explicit or not – helped us to craft the stories to further reflect the young people’s consciousness. We attended to ‘unexpected vibrations in unexpected places’ (Dimock, 1997), but also to no vibrations in expected places as part of the co-construction process. If we equate resonance with both ‘validity’ and ‘trustworthiness’ as Leavy (2016: 79) does, then we can claim some validity for our story-making. At a conference where the presentation of an animation about a child with autism was led by the young researcher whose story inspired it, a woman in the audience came to us at the end in tears provoked by resonance with her own daughter’s experience. Resonance in the wider audience – which we hope will occur in further unexpected and unpredictable places as we set the stories free – is an aspect of dissemination of the product.

Conclusions

Focusing on some of the details of the process of co-constructing stories with young people, academics, students and community writers, we have exposed the intricate processes of meaning-making in participatory research. Meanings are made in moments of connection between the different people involved, occurring in layers as stories are told, withheld, drafted, elaborated, edited and shared. Our analysis has shown that resonance during the process of story-making has the potential to enhance resonance with the final products.

We have drawn on a kind of collective narrative practice to challenge established discourses and narratives. To this extent, we have met our aim of creating stories that connect people and challenge stereotypes. While one might argue that distilling lives into stories and providing happy or ambiguous endings is a dilution and distortion of ‘truths’, we might also argue that ultimately we are revealing an activist position: we are not just telling ‘what it’s fucking like’, but ‘what it could be like’ – ‘for Others, for Us’ (Harrison et al., 2001: 339–340), with an important recognition that ARs and YRs are not much different. We are not presenting our stories solely as our findings but hope they will operate as well-informed resources that might help other young people to navigate issues that our research participants have identified. Therefore, the authorial claim for the text (Richardson and St Pierre, 2005: 961) is neither as pure social science nor as fiction, but as story.

We have argued that our collaboration with young people in this research has led to a more nuanced understanding of their lives, while also recognising that the ghosts of the young people and of ourselves are entwined in the making of meaning. Our reasons for making the stories are not simply for aesthetic approval or as a ‘writer’s personal signature’ (Leavy, 2016), but as a means ‘by which those truths which cannot otherwise be told, are uncovered’ (Clough, 2002: 8 in Watson, 2011: 404), not individually, but together.

Our stories acknowledge but counter the narratives of negative experiences, constrained expectations and limited perspectives that are frequently given about marginalised young people in the media and in literature. Rather than ‘looking away’ (Marlow, 2018) from disability and disadvantage, our stories are told from the inside, reflecting a truth imbued with humanity and hope. Now we hope that our experience (also distilled into a Practitioner Guide [Satchwell et al., 2018]) can help others to recognise the potential of story-making by, with and for young people to increase participation and connection.