Supported and valued? A survey of early career researchers’ experiences and perceptions of youth and adult involvement in mental health, self-harm and suicide research

The present study

Mental health research, including research into self-harm and suicide, is likely to present specific challenges to youth and adult involvement work (i.e. working with ‘vulnerable’ groups on topics deemed to be ‘sensitive’ in nature). The present study examines the perceptions of ECRs in these fields regarding youth and adult involvement. Specifically, we examine the experiences and perceptions of ECRs who have been involved in youth and/or adult involvement work in the field of mental health, including self-harm and suicide research, in order to: (i) determine the extent to which these ECRs feel that work in youth and adult involvement is valued, (ii) describe how ECRs in mental health involve young people and adults in research, and (iii) identify common barriers and facilitators in undertaking youth and adult involvement in their research. We hope to provide valuable and practical information about current practice, barriers and facilitators, with the aim of finding ways to better support youth and adult involvement work in mental health research.

Participants

The survey was open to ECRs undertaking research in mental health, self-harm and/or suicide in the UK (a convenience sample). A total of 41 participants completed the survey. Due to the small research field and familiarity of those working in this area, no additional identifiable information was collected (e.g. age, gender, current institution). Status as an ECR was self-determined and thus included postdoctoral and postgraduate researchers (amongst others).

Design and procedure

The anonymous self-report questionnaires were administered online. Participants were recruited via e-mail listings, a link on a website for ECRs (Early Career Researchers Youth And Public engagement on Self-harm Network) and via social media. As such, the total number of potential participants reached is unknown. The survey was made available in February and March 2017. Ethical approval was given by the departmental Research Ethics Committee (January 2017).

Youth involvement in this study

‘Youth Speak’ were a mental health research advisory group operating between 2013 and 2017 at the lead author’s institution (at the time of data collection). The members were aged between 14 and 24 years old and had contributed to a number of mental health research projects over the 4 years. The group was established upon the principle of developing a youth culture that promoted inclusiveness and collaboration, and as such the members were not asked about, or expected to disclose, their mental health difficulties.

The Youth Speak group were involved in the planning and content of a symposium for ECRs on youth and adult involvement in mental health research. The plans for the symposium were discussed at a Youth Speak meeting. Due to timing and financial constraints, the group members were not able to attend the symposium. Instead, we produced a short film presentation of the young people talking about their experiences of research involvement and engagement activities that was shown at the symposium. Ideally, it would have been advantageous for the Youth Speak group members to attend the symposium, particularly as they considered teaching researchers (including ECRs) how to work effectively with young people to be an important part of their Youth Speak role. Informal feedback regarding the symposium suggested that the ECRs would also have welcomed the opportunity to meet and discuss research involvement with the Youth Speak group members.

Measures

We did not provide respondents with definitions of any of the terms used in the questions. For instance we did not define what was meant by the word ‘value’ in the questions asking the extent to which the respondents/their organisation/their department value youth/adult involvement. We anticipate that our respondents understood the use of the term ‘value’ as meaning ‘held in high regard’/‘seen of worthwhile pursuit’, although it is possible that other interpretations (e.g. regarding monetary worth, or principles and beliefs held) were applied.

Data analysis

Quantitative descriptive statistics (analysed using SPSS V24 for Windows) are presented. Responses to open-ended questions concerning barriers and facilitators were analysed thematically using inductive coding. The coding mapped onto the questions asked, so that distinct ‘barriers’ themes and ‘facilitators’ themes were generated. The initial analysis was done by AJW and KB, with codes being reviewed and developed further by EN and RW, before the final coding was refined collectively by the group. To establish the reliability of the initial coding, 17 unlabelled extracts were presented alongside the theme descriptions to an independent coder, PR, who was blind to the original analysis and coded on their own. PR’s coding was then compared to the original coding, with good reliability (the independent coder attributed 82.4% of the extracts to the same codes as the original coders). Following this, a final framework was synthesised from the previous themes, which was reviewed by all authors to ensure clarity.

Youth and adult involvement in Research

Participants’ research interests were reflected in their research involvement work: this was primarily concerned with mental health, then suicide, self-harm and ‘other’ (see Table 2). The majority of participants had a supervisor or line manager who was also engaged in youth or adult involvement research activities. Similarly, most participants reported that their institution undertook such work.

The reported frequency (‘never’, ‘rarely’, ‘sometimes’, ‘often’, ‘always’) with which young people and adults were involved in the respondents’ research and their institutions’ research is compared in Fig. 1. Adult involvement was reported more frequently than youth involvement within the ECR’s own work, whereas these values were more similar for institutions.

Analysing data was the least utilised method of both youth and adult involvement. Across both populations, developing resources was the most popular method. Additional methods reported (in an ‘other’ response option) were involving people in developing or giving feedback on research tools, and research involvement groups training other people in research involvement skills.

Funding

Less than a third of participants (n = 13, 31.7%) reported that funding was available for their research involvement activities. Sources included research councils/funding bodies (e.g. as part of doctoral grants) or internal university funds. Thus, despite limited funding, ECRs managed to conduct research involvement (with only some utilising a variety of funding sources).

Barriers and facilitators

A total of five main themes describing challenges (three themes) and facilitators (two themes) to youth and adult involvement work were identified in the open-ended question responses. Two of these themes contained subthemes. The themes and subthemes identified are described below with example quotes.

The challenge of recruitment

Several responses indicated that difficulties in recruiting young people and adults were a barrier to involvement work. Some responses reflected difficulties in recruitment generally in the field “Recruitment of participants” (ID:34), “Finding the relevant people” (ID:16), “No links with societies, charities etc, therefore difficult recruitment …” (ID:7), “Reaching a representative enough group” (ID:29). Others highlighted challenges regarding working with specific populations “… I work with a population who by definition are ill, and so their own illness trajectory can get in the way for [these] individuals” (ID:9).

Expert examples, expertise and guidelines

Several respondents indicated that they would find it easier to conduct involvement work if they had the opportunity to learn from others with more experience: “Collaborations with experienced colleagues” (ID:7); “Access to learn from more experienced researchers about how they raise awareness of the opportunity for involvement, and the different ways in which the [y] involved people in their work outside of the traditional, steering group type set-up, which for me doesn’t quite cut it.” (ID:17). The merits of professional resources that could be used, such as training schemes and research involvement guidelines, were also highlighted “Having a framework or model to base engagement on.” (ID:15), “Possibly more training on engaging people on a sensitive topic such as suicide” (ID:8).

Perceived value of youth and adult involvement in research

ECRs considered youth and adult involvement in research to be valuable generally, as well as being valuable to their own research. They also reported that their institution and department appeared to value youth and adult involvement in research. However, a notable proportion of ECRs reported that their institution did not undertake involvement activities (or were not aware of such work). Ratings suggest that ECRs feel relatively well supported, confident and knowledgeable when undertaking youth and adult involvement in research. Although some ratings were lower for youth versus adult involvement, the only rating difference of significance was the perception of value that institutions gave to youth versus adult involvement (lower for youth). Youth involvement yields the same barriers as adult involvement, but may have additional barriers such as gatekeeping, ethical concerns about child involvement, and being considered a ‘difficult to reach’ population in general. It is also possible that young people are regarded as being less capable of making a valuable contribution to research, but this requires further exploration. It should be noted that most respondents had a supervisor or line manager who undertook youth and/or adult involvement, thus providing a potential source of expertise and support.

How ECRs involve young people and adults in research on mental health, self-harm and suicide

Both young people and adults were involved in research through similar methods across the research life cycle. Both groups were most frequently reported to be involved in developing research resources (such as participant information sheets) and dissemination of research findings. These are relatively discrete and time-limited activities within the lifespan of a research project (and arguably easier to manage and cheaper). The least often reported involvement activities for young people and adults were data analysis, preparing research applications and reporting the research (a more in-depth level of involvement along the co-production continuum) - tasks which possibly require more expert knowledge of research topics and academic processes. Youth and adult involvement in the analysis or reporting of research findings (for example, respondent checking and validation of themes in qualitative work) may not always be achievable or desirable, depending on the methodology. Indeed, this specific method might have a limited capacity within youth involvement (e.g. due to complexity of tasks and lack of research experience), however, arguably it is the responsibility of the researcher to make such activities accessible to the target audience. Public involvement in preparing research applications should arguably be more common practice given funding body PPI requirements, but it should be acknowledged that ECRs may not have been involved with their current research at such an early stage. Finally, compared to adults, young people were reported to be less frequently involved in identifying research questions, contributing to study design and in managing the research (in the form of an advisory committee, for example) - although our data do not speak to why this might be. If young people are not being involved in setting research questions and priorities, their views risk being ignored in the national youth mental health research agenda [12]. It might be that researchers are less inclined to engage with youth involvement due to the additional steps needed to adjust tasks to meet the abilities, preferences or language needs of younger people.

A previous analysis of the types of involvement activities reported in applications to the National Research Ethics Service (NRES) also found that dissemination of research findings was the most frequently reported activity whereas analysis was the least frequently reported activity [3]. Young people involved in mental health research have expressed a desire to be involved in every stage of the research life-cycle [13] and at the far-end of the involvement/participation continuum children can be successfully supported as researchers in their own right through child-led research [18, 19]. Future research could explore ways of facilitating youth involvement in mental health research further, for example, identifying those parts of the research process in which young people can be involved and where youth involvement can lead to improvements [20].

Barriers and facilitators in undertaking youth and adult involvement in mental health, self-harm and suicide research

It is problematic if youth and adult involvement activity is not going ahead in mental health/self-harm/suicide research due to unwarranted ethical concerns, particularly as this is something that ECRs may not feel in a position to challenge or debate. It is important to emphasise that there is increasing research evidence that participating in research on self-harm and suicide does not have a detrimental effect on participant well-being. A recent meta-analysis of the impact of exposure to suicide-related content in research protocols indicates a reduction in suicidal ideation and behaviour after being exposed to suicide-related content or suicide assessment. Notably, this reduction was largest in young people [21] There is no reason to assume involvement in such research in an advisory capacity should be any different (though future research could usefully address this).

The second barrier identified relates to the practical costs (time and money) that undertaking involvement work can entail (balanced against the perceived benefits) and the third relates to the challenges involved in recruiting people to be involved in research. ECRs have a limited timeframe in which to complete their research and face many other competing research priorities. However, successful approaches to engaging young people in ‘sensitive’ involvement activities, for example in drug and alcohol research, requires a researcher to be flexible and young-person centred, and crucially takes time [17]. Nonetheless, within this context, the respondents highlighted potential ways in which to facilitate their youth and adult involvement work: expert examples, expertise and guidelines; and investment in involvement work (practical resources, especially money; wider community support; a supportive culture within institutions). In particular, funding needs were highlighted by ECRs – indeed, over two thirds of ECRs did not have access to funding to support their involvement activities. INVOLVE offer guidance regarding payment for PPI group members to compensate them for their time and expertise. It would be difficult for ECRs to adhere to these best practice guidelines without access to sufficient funds (e.g., where funding is not available, involvement activities may be reliant on people giving their time for free. Additionally, ECRs may be unable to cover travel costs incurred by involvement group member). This potentially magnifies already complex power imbalances, if some people (i.e. researchers) are being paid for their time and expertise, whilst others (i.e. PPI members) are not. The costs of involvement have previously been cited as a barrier to encouraging user involvement in mental health services [22], but there is evidence that the provision of small research development bursaries for PPI can support involvement activities and help develop this aspect of research [23]. However, much of the other support required could actually be provided at minimal cost to institutions. There is ample published guidance on PPI, but perhaps more hands-on mentoring or discipline specific guidance is required, or even simply improved signposting to such resources. Some of this could be achieved by raising the profile of research involvement as a valuable research activity within higher education research – although this would require significant buy-in from institutions.

At a more individual level we would urge ECRs to think creatively about how to best achieve their research involvement aims, by utilising technology and social media (i.e. virtual involvement) and by networking and collaborating with research colleagues in their field(s). In order to act as drivers for change, ideally ECRs need the support of their supervisors, senior colleagues and institutions. There is an apparent lack of clarity regarding ethical and risk governance issues surrounding involvement work in mental health, self-harm and suicide research, which may pose a substantial barrier to youth and adult involvement in this research field. We would argue there is a need for more targeted guidance for involvement work in this area and improved access to (modest) financial support.

Limitations

The survey respondents constitute a highly selective and small sample, with an interest in youth and adult involvement work. Due to our recruitment strategy (social media) we do not know how many ECRs in mental health research chose not to participate. Future survey work could aim to recruit a larger sample who are not necessarily as invested in involvement work.

The terminology used to describe involvement work is confusing and used interchangeably (e.g. engagement, involvement, knowledge exchange, PPI). Although we attempted to be clear in our participant recruitment material and wording of questions, potential participants may not have considered themselves eligible to participate due to another term being used in their experience, or may not perceive the work they do to be involvement work (e.g. consulting an advisory committee). Again, the distinction between youth and adult involvement can also be variable and thereby impact how the two populations are considered.