Doing It Write: Representation and Responsibility in Writing Up Participatory Research Involving Young People

This article adopts a reflexive stance as the authors look back on their doctoral research projects; the first author exploring young people’s relationships with community radio, and the second author studying young people’s alcohol consumption practices and experiences, both in the North West of England, UK. The authors discuss the methods of data collection they employed, which enabled young people the opportunity to participate in meaningful ways. However, drawing on snapshots from their PhD theses, the authors question whether decisions made when writing up related to protecting anonymity, (re)presenting speech characteristics, and editing, independently of participants, potentially undid some of the hard work exerted in creating an equitable space for young people’s contributions, resultantly perpetuating the regulation of young people and keeping them ‘in their place’. The authors propose some recommendations for facilitating the inclusion of young people in the writing up of participatory research.


Introduction
Whilst I worked incredibly closely with the women at Sure Start Parr (as well as local artists and professionals who provided crucial assistance throughout the research project), I am aware now that, in writing this article, I am alone. Whilst I draw on our shared memories, this work is ultimately my creation. I acknowledge the power that I hold as I write about this project, and as I attempt to do justice to the extensive work to which we all contributed. (Foster, 2007, p. 368) Participatory research, originating from Tanzania in the 1970s (Hall, 2005), is ingrained in work with marginalised and oppressed people living in developing areas. Par-ticipatory research has since been studied and undertaken by scholars in disciplines ranging from Anthropology to Health. Child rights advocates, critical educators and youth workers have embraced the ethos of participatory research as research 'with' as opposed to 'on' participants. Participatory research is celebrated as actively involving participants in: data gathering (Gallagher, 2008); analysis (Morrow & Richards, 1996); dissemination (Pain, 2004); and follow-up action (Cahill, Sultana, & Pain, 2007). In reality, participants are most often involved in data collection, less so analysis (Mauthner & Doucet, 1998;Pain, 2004), and less frequently still dissemination (Mohan, 1999). Although we are not suggesting that participatory research is "parachute research" (Minkler, 2014, p. 245), participants are seldom invited to participate in research after data collection. When participants are involved in dissemination this is typically through presentations. Academics often get on with their job and go solo when writing up. The epilogue opening this article by Foster (2007), a researcher undertaking a participatory project at Sure Start Parr, a children's day care service, epitomises the power and responsibility held by the researcher when writing publications. Our article carves out a space for thinking about the inclusion of participants in writing up, beyond the potentially burdensome and time-consuming process of co-authorship. Instead, we think of ways we can use the power and authority we currently possess in writing up to include participants, to more accurately represent them, and to "do justice" to their contributions (Foster, 2007, p. 368).
Representation is a complicated issue in social research insomuch as the researcher reflects a vision of the reality of a participant that has been subject to distortion (Foster, 2009). With this article, we further critical discussions of participatory research (Cooke & Kothari, 2001;Gallacher & Gallagher, 2008;Mohan, 1999;Pain & Francis, 2004), recognising that representing participants in academic writing requires more thought and consideration in research that claims to be participatory. We reflect on our doctoral research projects: Connecting Communities through Youth-Led Radio, which explored young people's relationships with community radio (see C. Wilkinson, 2015), and Young People, Alcohol and Urban Life, which studied young people's alcohol consumption practices and experiences (see S. Wilkinson, 2015a). Although these projects were framed with "pockets of participation" (Franks, 2011, p. 15), we wrote up our theses independently of our participants. Herein, we critically reflect on how decisions we made when writing up ultimately (re)presented our participants.
First, we situate participatory research with children and young people in debates on writing up. We then offer a synopsis of our research projects (C. Wilkinson, 2015;S. Wilkinson, 2015a). Following this, we provide empirical examples from our research to reflect on our representations of participants, focussing on three areas: protecting anonymity, (re)presenting speech characteristics (such as accents and impediments), and editing decisions (for example the inclusion/exclusion of expletives in participant quotations). We conclude by emphasising the importance of involving young people in writing up participatory research, and provide some recommendations for how this can be achieved.

The Unwritten: Participatory Research with Children and Young People
The emergence of the sociology of childhood, which debunked the view of children as incompetent and "becoming-adults" (Lee, 2001, p. xii), has contributed to a reassessment of the inclusion and role of children in research. Participatory research has been positioned as one way to achieve this inclusion, supported by scholars who believe that, through involvement in research, people have a better opportunity to influence decisions concerning their lives (e.g. Crivello, Camfield, & Woodhead, 2009;Grasser, Schunko, & Vogl, 2016). Whereas children and young people have been, and can still be, considered marginal in research, participatory research positions them as co-creators of knowledge. In participatory research, children and young people are often employed as peer researchers because they are believed to possess skills that adult researchers do not: they speak the same language as their peers; they have access/membership to hard-to-reach groups; and they have first-hand insight into matters affecting other children/young people (McCartan, Schubotz, & Murphy, 2012). In this sense, they are experts in their own lives (Burke, 2005;Mason & Danby, 2011). Owing to this 'expert' insight, knowledge produced from participatory research with children and young people can be considered more authentic (Grover, 2004), richer, and more reliable than that produced through traditional top-down approaches.
By involving children and young people in research, they arguably "cease being data mules in the carriage of other people's academic careers" (Smyth & McInerney, 2013, pp. 17-18), and are realised as agentic and competent actors in their own lifeworlds. However, Mohan (1999, p. 51) is concerned that "despite replacing a monologue with polyphony there are still questions of who writes up, who publishes the material and whose career benefits?" Mohan (1999) reflects that young people are often not invited to participate in research post data gathering. Discussing a project that attempted to engage young people with an intellectual disability in participatory research, Dorozenko, Bishop and Roberts (2016, p. 200) argue that, as academic researchers, they had "certain skills and expertise that lent itself to research", such as undertaking literature reviews; analysing qualitative data; and publishing. Thus, it would be "self-effacing (and dishonest)" to deny their contributions to these stages of the project (Dorozenko et al., 2016, p. 200). Muhammad et al. (2015) support this, stating that academics have the training and expectations to produce peer-reviewed articles, whereas young people may have distinct responsibilities (for instance school/work) that preclude additional tasks. A potential consequence is that "academic power and privilege can become omnipresent" in the writing and representation of data (Muhammad et al., 2015(Muhammad et al., , p. 1055. We commend Mary Kellett's decision to include Ruth, Naomi and Simon, aged 10, as co-authors on an article about empowering children as active researchers (see Kellett, Forrest, Dent, & Ward, 2004). Kellett enables these children to take ownership of their research agendas, and challenges the status quo. Other authors (e.g. Cahill et al., 2004;Townsend et al., 1995) have been successful in producing publications collaboratively with participants.
Co-authorship may not be practical or desirable for research participants who have their own busy lives, and may also not be practical for the research project or researcher. There is a balance here then between participatory ethos and pragmatic decisions around cost, time and resources. There have been (justified) concerns of over-burdening participants in participatory research (see Cornwall & Jewkes, 1995;Flicker, 2008). However, some participants may wish to participate more fully than the remit outlined by researchers allows, and may become sceptical after being prohibited from participating in the ways that they expected (Barreteau, Bots, & Daniell, 2010). This relates to an important critique, that some participatory research projects involve young people in tokenistic ways, resulting in low levels of self-advocacy and empowerment (Cooke & Kothari, 2001;Mohan, 1999). Related to this, Thomson (2007, p. 207) highlights how participatory approaches can cause harm if children's perspectives are rendered meaningless, due to unacknowledged personal assumptions of the researcher which resultantly keep children "in their place". Gristy (2015) highlights the problems of representation and speaking for others within participatory research. In this view, representation in the communication of findings is a political act. This is especially so when considering editorial decisions, for instance the choice of publication venue; word limits; and restrictions to lengths of quotations etcetera. Following Ansell (2001), choices must be made by the researcher, and although the consequences of these choices can neither be fully controlled, nor fully known, some responsibility must be assumed for the potential outcomes. With this article, we pinpoint aspects of writing up where further dialogue and joint decision-making is needed between academics and participants to do justice to participants' contributions, and to represent participants in ways they are happy with.

Connecting Communities through Youth-Led Radio
C. Wilkinson's (2015) research project Connecting Communities through Youth-Led Radio explored the ways in which KCC Live, a youth-led community radio station in Knowsley, neighbouring Liverpool, UK, provides a space for young people to find and realise their voices, build stocks of social capital, and create their own communities. KCC Live was founded in 2003 as a college enrichment and work experience radio station, based at Knowsley Community College. KCC Live acts as an element of the college's retention strategy and intends to function as a bridge for young people Not in Education, Employment or Training (NEET) to (re)enter the labour market, though not all volunteers are NEET. The station typically has a 14-25 year-old volunteer base.
This research project adopted a participatory design in collaboration with 21 volunteers/staff members at KCC Live. Mixed methods were employed, including: 18 months of observant participation; 1 interviews and focus groups with volunteers; interviews with management at KCC Live and Knowsley Community College; a listener survey, listener diaries, and follow-up interviews. The implementation of some of these methods contained participatory elements. For example, the young people were involved in designing and refining interview questions for management through mind mapping sessions. The young people and the researcher also co-produced the listener survey, and the young people assisted with distributing the survey. Accompanying the thesis were two co-produced audio artefacts: an audio documentary, 'Community to me is…', which explored young people's understandings of community, and a three-part radio series, 'What we found', in which the young people discussed the research findings. The young people assisted in the recording and editing of these audio artefacts. Despite participating at various stages of the researchincluding the audio dissemination-the young people were not invited to participate in writing up.

Young People, Alcohol and Urban Life
S. Wilkinson's (2015a) research project Young People, Alcohol and Urban Life explored the alcohol consumption practices and experiences of 40 young people, aged 15-24, living in the suburban case study locations of Wythenshawe and Chorlton, Manchester, UK (see also S. Wilkinson, 2015b). This research was conducted with young people, using a flexible suite of methods which they could 'opt into' (Leyshon, 2002, p. 182, emphasis in original), including: interviews; peer interviews; drawing elicitation interviews; diaries; mobile phone methods (S. Wilkinson, 2016); and participant observation. Offering a palette of methods enabled participants with different skills to participate in ways that were meaningful to them. The author refined and developed her methods through listening to the preferences of participants. For instance, some young people asked if they could be interviewed with their friends in what the author labelled 'friendship group interviews'. This illustrates the agency of participants to shape the research design.
Participants were also given the opportunity to interview friends about alcohol consumption. The peer interview method is a participatory tool that provides a way of foregrounding the perspectives of young people (Kellett et al., 2004). The researcher ran informal interview training sessions for the young people. This training equipped the young people with new skills, such as designing an interview schedule and gaining consent, which gave them more control over the project (see also Schäfer & Yarwood, 2008). Further, the transferable skills (Lushey & Munro, 2015) gained by young people, such as asking powerful questions and listening carefully, could be useful when seeking employment. Whilst some young people were more content using interview questions designed by the researcher, others were keen to develop their own questions (see also Schäfer & Yarwood, 2008).
Analysis of data was undertaken independently of young people in this study, as was writing up.

Reflections on Writing Up
Our individual research projects had much in common: they were both ethnographic, undertaken when we were doctoral students, employed multiple methods, and engaged with young people in data collection but not writing up. However, it is worth emphasising that the different topics (community radio and alcohol) may highlight different considerations in terms of writing up; for instance, aspects related to the representation of speech and voice may be given more of an emphasis for those involved in community radio, whereas concerns over anonymity may, arguably, be more pressing for a study concerning alcohol consumption. Given the young people's non-participation in the writing up of our theses, we now unpack aspects of our writing which required us to (re)present participants, considering three key areas, respectively: protecting anonymity; (re)presenting speech characteristics; and editing decisions.

Protecting Anonymity
An important part of ethical practice in most research is ensuring participants' anonymity (Grinyer, 2009). However, in both of our research projects, some participants expressed a desire to be named in our theses and future publications (see also Pymer, 2011). In the second author's research, one young person questioned: "are we going to be famous?" (Author's field diary, 15/11/13), demonstrating pride at appearing in published work. These young people wanted to showcase their involvement to others (Wiles, Coffey, Robinson, & Heath, 2012). However, we both decided that revealing names would compromise the anonymity of participants (Trell, Hoven, & Huigen, 2014), which may have negative future implications-for instance, when seeking employment (particularly in S. Wilkinson's 2015a study of alcohol consumption practices). It is important to stress that we, supported by our respective university ethics committees, made the decision of what was 'best' for our participants.
To give the young people more ownership over their stories, the first author allowed young people to choose their own pseudonyms. After a group discussion prompted one young person to suggest choosing pseudonyms after pop stars, DJs and presenters, others were enthused by this idea and proceeded to select aliases from their celebrity idols. Interestingly, many young people questioned "are we allowed?" and "would I be allowed to call myself that?" (Author's field diary, 23/07/13), perhaps illustrative of their views of academic outputs as serious and mundane. Within Moorefield-Lang's (2010) research with middle school students, some participants chose to name themselves after cartoon characters. Akin to Moorefield-Lang (2010), the first author believed that allowing young people to choose pseudonyms enhanced the participatory nature of the study, also affording the young people greater agency. As a result, young people featured in C. Wilkinson's (2015) thesis, related publications and conference outputs as: Madonna, Robbie, MJ and Modest Mouse, amongst others.
The second author also planned to allow participants to choose pseudonyms, to protect them from feeling a loss of ownership over their work (Alderson & Morrow, 2011). However, she later decided against this and allocated each participant a name. This was because, where members of the same family participated in the research, family links may be more easily guessed; for instance, by inadvertently using the real name of a relative, friend, or associate (Pymer, 2011). However, looking back, the author questions the appropriateness of some of the names she allocated to participants. For instance, she named one participant Vera, an arguably dated name, when the participant's real name was much more modern. Grinyer (2009, pp. 49, 52) also reflects on the conundrum of choosing "equivalent" names for participants, and the "unanticipated distress" caused by allocating names that have negative associations or that participants cannot relate to. In summary, as researchers we possess a certain level of power and authority in writing up and in choosing pseudonyms for our participants. We argue that anonymity needs to be discussed with each participant on an individual basis to ensure they have the level of ownership they desire over their own spoken words.

(Re)presenting Speech Characteristics
Translating the spoken words of participants into text, a "static form of representation" (Mero-Jaffe, 2011, p. 232), requires a number of often taken-for-granted decisions by the researcher (or external transcriber). In our projects, we undertook the transcription ourselves as we wanted to retain closeness to the data. Accents, predominantly Liverpool (Scouse) accents, were a prominent part of data gathering in Connecting Communities through Youth-led Radio (C. Wilkinson, 2015). The author transcribed the data verbatim, attempting to accommodate nuances in the accent, see the excerpt below: The only difficulties I've faced is getting into like doing me show properly and like doing me voices and things. (Fearne,22,interview) Above, the word 'me' used by Fearne in place of 'my' is characteristic of the Scouse discourse. In capturing this, the author attempted to honour Fearne's discourse and accent, considering such nuances "sonic gems" (Oleksik & Brown, 2008, p. 163) in a study concerned with youth voice. However, there is evidence (see Corden & Sainsbury, 2006) that regional expressions present in verbatim quotations can be considered unattractive, and may lead to negative judgements about the speaker.
Another instance which required the first author to think carefully about how to (re)present speech concerned a participant with a stutter. The author decided not to make the participant's stutter evident in the transcribed data and therefore omitted pauses and hesitations, as well as repeated words, sounds and syllables. The author made this decision due to awareness that quoting things as they appear can be hurtful to participants (Beason, 2000). Further, as only one volunteer at KCC Live had a speech impediment, reflecting this in written text may have compromised the participant's anonymity, particularly when combined with other information disclosed in the data (see Oliver, Serovich, & Mason, 2005). However, this is complicated as the "rule of thumb" is that data should be presented in such a way that participants can recognise themselves, while the reader cannot identify them (Barnes, 1979, p. 39). The author was concerned that the participant may not recognise himself in her (re)presentation of his speech. Demonstrating this conflictual issue, some participants in Corden and Sainsbury's (2006) study believe that attention should not be drawn to impaired speech, whilst others were concerned about the participants' reactions to knowing the researcher had changed their speech. Notably, the first author was reluctant to omit the stutter because she believed it had significance to the focus of the research on voice. Similar considerations were not as marked in the second author's study of alcohol consumption.
Following Mero-Jaffe (2011), the transfer of transcripts to participants is useful for validating the data, preserving research ethics, and empowering participants by allowing them control over what is written. We add here that the transfer of transcripts to participants should also be concerned with how things are written. The first author sent transcripts to participants to ask them if there was anything they were 'unhappy with' or would like to be removed. No participant requested changes to their transcripts regarding content and 'what' they said. One participant replied to say "amazing how many erms I say. I sound like a right tool box"-when the author asked if the participant would like her to remove the 'erms' the participant insisted "no! I just thought it was funny" (Author's field diary, 21/02/2015). Mero-Jaffe (2011) reflects how the minute detail of verbatim transcriptions has the potential to insult interviewees, who might feel that "natural features of their talk" (Pinter & Zandian, 2015, p. 242) were unrefined. Further, while certain authors reflect that it can be embarrassing for participants to read transcripts (Forbat & Henderson, 2005; Mero-Jaffe, 2011), we argue that it is important that this potential embarrassment occurs at the reviewing transcript stage, when it is possible to make changes to address any dissatisfaction, as opposed to seeing extracts from transcripts in published work. Thus, to ensure an equitable space in writing up, participants' views and feelings on the transcript must lead the researcher's decision-making about how to present data.

Editing Decisions
Editing decisions made by the author (and sometimes suggested by reviewers) can play a part in (re)presenting research participants. Mauthner and Doucet (1998, p. 138) reflect that: "we dissect, cut up, distil and reduce their [participants'] accounts, thereby losing much of the complexity, subtleties and depth of their narratives". Editing can be considered a "balancing act" (Pymer, 2011, p. 197) between removing certain information, whilst ensuring data remain meaningful.
One editing decision we were faced with was whether to include/exclude profanity. Within both of our data collection experiences, we were met with instances of participants swearing. The first author made a decision to exclude these swear words, using ellipses to indicate a word had been omitted, believing that including profanity would lower the tone of her doctoral thesis. As Foster (2009, p. 234) tells: "much of the life and exuberance of the research process may, necessarily, be omitted or flattened in order to produce an acceptably academic account". The second author included expletives within her thesis: Going to have a fucking crazy one tonight at Deansgate. 2 Been waiting for this all fucking week man! (John, 22, Wythenshawe, text message) Some people were sluts and went with lads that they would never even look at if they were sober. (Jenny, 16, Wythenshawe, drawing elicitation interview) Whilst being mindful that the presence of 'bad language' in academic papers can result in messy texts (Barker & Weller, 2003;Kvale, 1996), S. Wilkinson (2015a) believed that removing the swear words would be stripping emotion and affect out of the writing. By quoting participants extensively verbatim, instead of solely representing the key message, she hoped to offer a more authentic representation of the young people's views. Some participants in Corden and Sainsbury's (2006, pp. 105-106) study believed that removing profanity would result in a publication that is "untrue", whilst others stated that reading swear words would give the impression of the participant as "ignorant" and "not very nice". This illustrates the extent to which both of our decisions, made without consulting participants, potentially influenced a reader's perception of them.
We were also tasked with deciding whether to correct the young people's English. Neither of us corrected young people's English where it was grammatically incorrect, and did not to use the term 'sic' to indicate errors originating with the participants. Below is an excerpt from a participant's diary in the second author's study which shows both grammatical and spelling errors: It was a mans flat and there was about 60 people there. To be honest I was the only person that was drunk and everyone else was taking pills. The pills are called Nintendoes and they made you illusinaite like a Mario game. (Jemima,15,Wythenshawe,diary) Following Townshend and Roberts (2013), we believed that transcribing verbatim allows young people to use their voices and also avoids interrupting the narrative's flow. Proof-reading the above and correcting errors would have eradicated heterogeneity and potentially the different educational abilities of participants. However, it must be acknowledged that a participant may want their 'mistake' to be corrected, and may despair at seeing something grammatically, or even politically, incorrect attributed to them in published work (albeit by a pseudonym). Editing must be undertaken in the best interests of research participants, and this can be achieved in conversation with participants. By including young people in the writing up phase, we are more likely to sustain the equitable space carved out for young people as is currently active in data collection and, to a certain extent, data analysis.

Conclusion and Recommendations
It is often assumed that participatory research is a positive ethical and political framework for research with children and young people. Indeed, it can be, but this article has highlighted that the current rhetoric of participation risks setting up norms of appropriate engagement by implying that children and young people should participate in certain ways and not others (Gallacher & Gallagher, 2008). Despite advances in researching 'with' rather than 'on' participants, this article has identified that the writing up of research remains primarily the responsibility of the academic researcher. We have presented a number of ways in which this can be problematic, owing to decisions made by the academic, independent of participants, which ultimately affects their representation. It is important to stress that we, as researchers, made the decision of what was 'best' for our participants. This highlights wider ethical concerns outside of academia, such as media representations of the lives of young people, and notably the demonisation of teenagers.
With this article, we argue that including young people in the writing up of research is important as, through "owning and controlling" participants' stories, researchers can sustain hegemonic depictions of participants, and worse still, add further to this oppression (Lynch, 2000, p. 80). In our research projects, we were guilty of this by denying our participants' desires to be named in publications (C. Wilkinson, 2015;S. Wilkinson, 2015a); attempting to replicate speech characteristics in arguably demeaning ways (C. Wilkinson, 2015), and mak-ing the decision to omit (C. Wilkinson, 2015) or include (S. Wilkinson, 2015a) profanity without consulting our participants. Even though work has been done to create a more equitable space for young people in participatory research agendas, adult researchers still maintain authority and power when writing up. When full-scale participatory research is not possible or desired, participation can always be improved (Greenwood, Whyte, & Harkavy, 1993). Thus, it is all about finding "appropriate and desirable levels of involvement" (Flicker, 2008, p. 84), without burdening participants or diverting them from other duties in their lives.
There are issues related to temporality, as the process of publishing an article can take months or years. As such, involvement-and particularly sustained involvement-from participants throughout this process may not be viable. We offer some recommendations for ways to facilitate the meaningful inclusion of children and young people in the writing up of research that do not require their full participation up to the point of publication. First, it is important to build participation into the writing activity. Second, we encourage collaborative publications, including the names of project partners as co-authors. We suggest that these recommendations can be pursued in the following ways: develop participatory guidelines pertaining to ownership, authorship, and dissemination; show participants verbatim transcripts of their data and encourage them to make suggestions and amendments, as opposed to asking them if there is anything they are 'unhappy with'; share drafts of papers/chapters with participants for feedback and be prepared to include dissenting views if there is disagreement on interpretation; constantly question whose voice is dominant in written work, and whose language is privileged. For these recommendations to be participatory, a trusting and mutually respectful research relationship is required.
Importantly, research projects often have limited funding which could prevent a researcher from revisiting the field after data collection and in advance of publication. Thus, with this article, we make a case for making this type of involvement part of the funding bid, and urge researchers to consider allocating a portion of funding to activity such as that recommended above. In line with Greenwood et al. (1993), researchers must continually evaluate the ways in which different facets of their research serve to enhance participation in order to avoid keeping children and young people "in their place" (Thomson, 2007, p. 207). Following the above recommendations, we believe that participatory research with children and young people can be more considerate to the ways in which they would like to be perceived in academic writing.