User involvement in the making: Positions and types of knowledge enacted in the interaction between service users and researchers in user panel meetings

Abstract Background Numerous studies of user involvement in research have been conducted. However, there is a lack of studies applying observational methods and addressing the concrete practice of involvement. Objective To determine what knowledge types and competences users apply when involved in the research process through user panel meetings. Design User panel meetings in a qualitative project in rehabilitation were sound‐recorded and transcribed verbatim. Data analysis applied an abductive approach framed by positioning theory. Setting and participants Six rehabilitation service users and a similar number of researchers met 20 times during a six‐year project period. They discussed various issues in the research process such as interview guides, analysis and dissemination of results. Findings The service users combined their respective knowledge and competence into six positions enacted in the panel interactions. They engaged as co‐researchers, based their contributions on their respective personal histories, represented an NGO and peers, applied their respective professional and educational backgrounds and, finally, engaged as concerned citizens. Discussion and conclusion The findings add to the discussion of professionalization of user involvement by introducing a wider array of positions enacted than do the findings of previous studies. Researchers recruiting user panel members, as well as NGOs appointing candidates for user panels, are advised to consider a wide competence profile for possible candidates. A panel is also considered as a resource in confirming and elaborating on a study's findings. Patient and public contribution A service user panel contributed to the study.


| INTRODUC TI ON
The last 20 years have seen the increasing involvement of patients in steering health research. Grant applications are now expected to outline how people whose lives the proposed research will affect are to be involved. 1 These studies typically identify the planning process and recruitment of participants as areas where user representatives have an impact. 4,5 Another group of studies addresses the users participating. One question is who is recruited to user participation. The studies addressing recruitment demonstrate that highly skilled, resourceful and strongly engaged people dominate in user involvement in research. [5][6][7] A second question concerns the impact on users and their social status when users are included as co-producers of research.
Their participation is judged to be an empowering process stimulating further activism on behalf of the group they represent. 8 A third question addresses how users participate in the research process.
Here, a key issue is the possible professionalization of users because they often turn into scientifically engaged lay experts. 7 In the critical disability movement and in mental health research, epistemological challenges to user involvement in research have been formulated. In an early contribution, Mike Oliver distinguished between positivist and interpretative science on the one hand and emancipatory science on the other. The latter is based on how disabled people formulate their interests in order to change society and fight ableism. 9 A recent contribution to emancipatory disability studies highlights the importance of disabled people playing an active, paid role in knowledge production and in the case of critical autism studies 'ensuring there is no sustainable dichotomy between autistic and non-autistic authorship'. 10 Psychiatric research is another arena where epistemological issues have been addressed. It is asserted by the psychiatric survivor movement that they have not fought to have a voice in research only to improve the research process as helping hands. Given the history of oppression in psychiatry, the power structures within the discipline need to be altered. [11][12][13] The current study does not fundamentally challenge the power relations in rehabilitation research but contributes to the recognition of competence by analysing the wide scope of knowledge enacted in user involvement.
In the studies of how users participate, two lacunae can be observed. First, what do user participants contribute when involved in the research process? This is a question seldom addressed in detail.
Second, studies based on observational data are scarce. The current study will contribute to lessen this knowledge gap. A user panel has All questions address role behaviour and will be well facilitated by a theoretical framework that engages with positions taken in social interaction.

| P OS ITI ONING THEORY
The position of selves is not static, but a dynamic process taking place in encounters between social actors. In a seminal paper, Davies and Harré outline a perspective where a concept of positions replaces the role concept. When people meet and interact, positions are created. 14 A key point is that social actors both relate to the ostensible topic at hand when meeting and bring in a diverse set of elements from their biographies: 'In speaking and acting from a position people are bringing to the particular situation their history as a subjective being, that is the history of one who has been in multiple positions and engaged in different forms of discourse'. 14 What goes on is a form of prepositioning whereby people bring their skills, character traits and personal biographical patterns of experience to the situation at hand. 15 In a discursive practice, of which the series of user panel meetings is an example, subject positions are made available to participants. These positions are built on both the conceptual repertoire a person possesses and the structure of rights to use this repertoire.
The panel meeting as a discursive practice is such a structure that actualizes specific concepts, experiences and story lines as meaningful when contributing to the discourse. Specific speakers are positioned by the ideas they bring into the conversation about their own positions. In addition, the conversation itself will contribute to the positioning of the specific speaker by the other participants and their ideas about that speaker. 15 In a user panel, participants will typically scan their pasts for occasions similar to the panel meeting so as to pick up hints about what is expected of them. The positioning of lay members of the panel will be further moulded by how a participant is understood by the other discourse participants, in this case other lay members and the researchers.

| ME THOD
The interdisciplinary research project Transitions in Rehabilitation, carried out at Oslo Metropolitan University in the years 2013-2019, established a user panel that met 3-4 times a year. 16 The study was planned to end in 2017 but the project period was extended because of two maternal leaves and one case of long-term illness among the PhD candidates contributing to the study. The project was a qualitative study of accident-injured people suffering from traumatic brain injury and multitrauma, and the professionals facilitating the rehabil-  presentations by researchers that users responded to and discussed among themselves and with the researchers. Most panel meetings were, in agreement with the participants' informed consent, sound-recorded and transcribed verbatim. Three meetings were documented by minutes only.
The data analysis followed the spiral approach outlined by Cresswell. 17  In the analysis, we applied an abductive approach. Unlike induction, abduction begins with data derived from a theory-informed research agenda, but unlike deductive analyses, abductive analyses are not bound by hypotheses formulated before the production of data. 19 Hence, abduction is a creative process aimed at producing knowledge and new hypotheses from unanticipated empirical observations. 20 Our unanticipated empirical observation was that lay panel members actively used their competences as professionals. We used positioning theory to organize this observation into a framework for further analysing the social interaction that occurred in the meetings.
The authors of this article are identified as R3, R4 and R5 in Table 1. Because we participated in the meetings, our analysis has an element of reflection on the process(es) in which we participated.
The idea to study positions and types of knowledge enacted emerged from our participation in the meetings. To ensure transparency, we have strictly kept to information represented in written transcripts of the meetings. Additionally, the abductive approach adds to the transparency of the analyses. The development of Table 2

| FINDING S
The user representatives enacted five social positions and based their engagement and contributions on six broad types of knowledge. Concerning knowledge learned from their personal histories, this knowledge can be divided into three aspects. These positions and types of knowledge are outlined in Table 2.
The positions and the knowledge types emerged in closely interrelated ways, and they were facilitated by the interactional patterns in the panel meetings. The presentation of findings will be organized around the five positions enacted.  interpretations suggested by researchers. The papers presented were based on interviews with both injured individuals and professionals involved in the traumatic brain injury (TBI) and multitrauma rehabilitation processes. At an early stage in the series of user panel meetings, an introduction to qualitative methods was agreed upon.

| The co-researcher
The initiative came from the user representatives, and in meeting 5, R6 and R7 carried out a teaching session on qualitative methods.
Users differed in their approaches to the analytical process. U4 held a master's degree in social science and frequently contributed analytical perspectives based on his educational background and his intimate knowledge of the logics of research in the social sciences.
In analysing professional work, the other users frequently were opposed to the analytical language applied. Below, R3 is referring to statements in a previous discussion about using analytical language.

| The affected individual
In Such reflections were instigated by both the evaluations of the participation process and the fact that the participation was to be studied.
The relations between the personal and representing peers were of course actualized by the fact that user participants were assigned to the panel by the NGO where they were active members.

| Disability NGO representative
U1 is the one who most explicitly brings in perspectives and priori-

| The professional
The study design included interviews with both professionals and

| The concerned citizen
The Another example was when gender as an analytical concept was introduced in a draft by R11 (who was not present in the panel meeting).
U5: I do not find it well motivated what she has presented. I do not find it valid. Men and women are different. We have different roles. We know that [laughs]. We are very different. That is not surprising.
These are the thoughts I have.
Here, U5 refers to her belief that gender role differences are manifest and legitimate. This belief leads her to problematize the critical analysis of gender roles determining how the injured prioritize in their rehabilitation processes.

| Distribution of positions
The various positions outlined were not evenly distributed among users. As already indicated, U1 most clearly formulated the NGO position, U4 the co-researcher position, U5 the professional position, U6 the engaged citizen position and U2 and U3 the most engaged use of their personal stories in order to elaborate and comment on findings. on users as co-researchers. 22 They concluded that this type of participation gains too much attention at the cost of research quality.

| D ISCUSS I ON
Forbat and Hubbard concluded that users as co-researchers conducting interviews and introducing their own experiences changed the topic of conversation in a way that was recognized as counterproductive to the research interview as information gathering. 23 Another challenge is the possibility that research designs will be forced in directions that favour the interests of specific patient groups. 24 The current study adds to this critical engagement by indicating a need for reflection on users' contributions to the analytical process. In the findings section, objections to the use of analytical  25 As demonstrated in the findings, one example was the panel's expanding on the importance of bowel and bladder functioning to disability discourse. These reflexions could have supported the analysis as additional empirical material but were difficult to include in the paper to be published on the experiences of the injured. A methodological requirement in qualitative research design is member checking, meaning that one or more representatives of the studied group are introduced to the analysis and contribute to validating results. 17 Such a procedure occurred in the panel meetings. An added value was the professional backgrounds in health and educational services held by two user participants.
They could contribute to member checking of the analysis of professional work from their positions as service providers. Based on these characteristics of the panel discussions, projects appointing user panels could consider the possibility of more actively including panel discussions as part of the methodological design.

| Limitations
The authors of this study participated in the social interactions

| CON CLUS ION
This study demonstrates the wide range of social positions and types of knowledge used by service users in a panel organized by a qualitative interdisciplinary research project in rehabilitation. Such a variety of competences is hinted at in other studies as well, but not systematized as in the current study. The systematization of positions is well suited to advise researchers recruiting user panel members as well as NGOs appointing candidates for user panels. The influence on the research process will not be based only on patient experiences and priorities set by a recruiting NGO, but also on the professional competences and educational training held by participants, as well as engagement in social issues on a broad scale.

| CONFLI C T INTERE S TS
The authors declare that they have no conflicting interests to disclose.

ACK N OWLED G EM ENTS
The authors thank the researchers and user participants involved in the Transitions project for valuable feedback on drafts of this article presented in user panel meetings.

DATA AVA I L A B I L I T Y S TAT E M E N T
The data that support the findings of this study are available upon reasonable request from the corresponding author. The data are not publicly available due to privacy restrictions.