Reporting participatory methods and author positionality in autism

In 2021, the journal Autism followed the lead of the British Medical Journal by introducing a mandatory Community Involvement Statement for all Research Reports. The Community Involvement statement was designed, as described in our manuscript submission guidelines, to encourage: ‘research that is actively carried out “with” or “by” members of the Autistic and autism communities (rather than “to”, “about”, or “for” them)’. A range of terms are used to describe this kind of work, including coproduction, participatory research and patient and public involvement and engagement (PPIE). Now, as the journal adjusts its author guidelines in relation to community involvement, we want to share the decision-making behind that change and reinforce our commitment to both doing and reporting participatory work.

Community involvement is desirable for a host of moral, ethical, scientific and practical reasons (Fletcher-Watson et al., 2019, 2021). Fundamentally, the involvement of autistic people in work that seeks to define them, map their experiences and abilities, or create and measure supports designed for their use, is a moral imperative. This aligns with the disability rights slogan nothing about us without us and even a moment’s reflection reveals how unreasonable it is to claim to be doing valuable and excellent autism research which takes no account of the views and first-person experiences of autistic people. Clearly, this is still an evolving practice however – fully-integrated and in-depth involvement takes skill, resource and time, and is out of reach for many small-scale research projects. But researchers can nearly always take time to reference autistic writing (Jaswal & Akhtar, 2019), draw on priority-setting exercises (Cage et al., 2024; Roche et al., 2021) or share ideas for feedback via social media, where many autistic people are active (Hudson et al., 2023).

Community involvement can also help to make work more ethical, meaning safer for participants and a better use of their time, when autistic people contribute to the design of methods and materials. Involving representatives of the participants being recruited can ensure that information sheets and consent forms are accessible, genuinely addressing participant questions. Working with autistic people can ensure maximum clarity about what to expect from a research study, given the known lack of precision and clarity in much writing by non-autistic people. Scientific benefits include advantages when it comes to recruitment, as endorsement from the community can shape more effective study adverts and help tap into community networks to circulate opportunities. Data quality and completeness can likewise be raised when measures, experiments or surveys have been co-designed for relevance and acceptability. On the practical front, any efforts to translate research insights into practice will be significantly advantaged if end users have been involved from the outset (Ellis & Nicholls, 2024).

This description of the imperatives for doing community-involved research brings us to the first limitation of our original community involvement statement – the lack of space or indeed author direction to report on methods, let alone their effects. The majority of authors reporting in this section provide a few sentences of content, at most, and a recent analysis of published papers in 2022 (the year after the statement was introduced) revealed that while reporting increased fivefold, 46% of those that provided a statement declared no community involvement (Tan et al., 2024). Moreover, many articles coded in this analysis were found to contain insufficient information on who was involved, and in which activities, to allow for categorisation. Thus, our existing approach to reporting community involvement could be improved, specifically by providing more detail on what reporting is expected and allowing additional word count to authors completing this section (Weschke et al., 2023). This second element is crucial. If we want authors to report what actually happened, in such a way that reviewers and readers can understand, critique and perhaps replicate the approach, we need to provide them with sufficient space to do so. This way, reporting methods of involvement goes beyond providing a platform to declare that involvement took place, and adds in an opportunity to educate less experienced readers in the techniques involved.

A second limitation to our community involvement statement has been uncovered via a number of routes. As highlighted in a recent editorial (Ellis & Nicholls, 2024), autistic people can and should be part of the research process as researchers, not only as representatives of a community which exists outside academia. We are fortunate that many autistic authors choose to send their work to Autism for publication. Some of these authors have reached out to us over the years since the compulsory community involvement statement was introduced, to note the difficult position in which they were placed. In a study in which no formal participatory methods were deployed with autistic people outside the research team, should they report that there was no community involvement, thus erasing their own autistic identity? Or publicly identify as autistic in a journal article? As one author expressed ‘I have not disclosed my diagnosis . . . and don’t feel a journal policy should be what prompts me to do so’. Research validates autistic people’s reluctance to disclose their diagnosis at work (Romualdez, Heasman, et al., 2021; Romualdez, Walker, & Remington, 2021) and the diagnostic disclosure apparently-demanded by a compulsory community involvement statement goes far further than disclosure to immediate colleagues. It remains a concern that there may be more autistic authors for whom this has become a problem, who do not want to disclose to our editors privately and are therefore unable to raise the issue.

Our initial response to this concern was to augment and clarify the author guidelines with a Frequently Asked Questions (FAQs) document, including providing a range of reporting options in situations where one or more authors identifies as autistic. However, these FAQs cannot provide for all situations. Moreover, we do not want our autistic colleagues to face an extra mental, emotional and administrative burden as they navigate this journal requirement – something that non-autistic authors do not have to do. As a journal, we want to encourage autistic research leadership, but we are forced to recognise that a compulsory community involvement statement adds a penalty selectively to autistic authors. As a result, we have become concerned that the compulsory nature of the community involvement statement is causing unintended harm to some of the most important stakeholders in our journal.

A third element of this deliberation concerns the role that personal characteristics and beliefs play in the research process. Another recent editorial (Milton & Green, 2024) discussed the importance of reflexivity in researchers and the role of often-hidden paradigmatic assumptions that drive underlying theory, research practice and interpretation. Despite the potential importance of situatedness and positionality in research, our journal currently offers no author guidelines on reporting in this respect. This intersects with the discussion above about participatory methods. While our goal is to remove any selective (and therefore ultimately – if unintentionally – discriminatory) burden from autistic contributors to the journal, we also know that some autistic authors have welcomed the space provided by the Community Involvement Statement to share their identity. Creating an optional space for positionality statements prompts all authors to reflect on the assumptions and experiences they bring to their work and share those.

As a result of these reflections, and with gratitude to many who have advised and informed the process to date, our author guidelines are changing from September 2024 onwards as follows:

We do not claim that these new author guidelines represent perfection. For example, the guidelines to some degree presume a boundary between community member and author, when in reality the former role can evolve into the latter, and autistic authors are always part of the autistic community. Nonetheless, we hope that the clear separation of reporting about methods from reporting about personal experience, with no compulsory expectation, will help people determine which sections are relevant for their manuscript. In the future, we expect these opportunities to share the ways in which different people influenced the research being reported will also be joined by introduction of CRediT statements ¹ – this is being rolled out more widely by the journal publisher over the course of the next 2 years, and at Autism, we look forward to adopting this systematic approach to reporting author contributions. Importantly, when the time comes, we want to ensure that co-authors without academic qualifications, for example, do not feel dissuaded from co-authorship by the need to specify their contribution in this way.

Our journal goal is to facilitate inclusion of diverse roles in autism research, respected and valued equally, and described and acknowledged appropriately. While an ideal world would not require any special guidance or nomenclature for specific groups of contributors, in autism research, we have a history of power and decision-making being dominated by non-autistic people, and of autistic contributions being diminished or invisible. It might take some proactive measures to overcome that. We hope our new Participatory Methods and Positionality Statement sections provide space to uplift autistic research contributions, and the contributions of people with other marginalised identities. At the same time, making these statements optional is an important way to avoid singling-out specific types of contributors for additional labour. We end by renewing our journal commitment to both motivate and report on diverse roles in autism research, to deliver work which has autistic benefit at its heart.