Epistemic justice in public involvement and engagement: Creating conditions for impact

Abstract Introduction Patient and public involvement in research is anchored in moral and epistemological rationales. Moral rationales relate to the public having a right to influence how knowledge about them is generated. Epistemological rationales relate to how research design and implementation can improve when informed by experiential, as well as technical, knowledge. In other words, public involvement can increase the epistemological resources of researchers, and contribute to research that is fit for purpose and has high external validity. Methods This article presents an analysis of 3 meetings and 11 interviews with public collaborators and researchers in three UK‐based health research studies. Data comprised transcripts of audio‐recorded research meetings and interviews with public collaborators and researchers. Data were first analysed to develop a data‐informed definition of experiential knowledge, then thematically to investigate how this experiential knowledge was considered and received within the research space. Results At meetings, public collaborators shared their experiential knowledge as stories, comments, questions, answers and when referring to their own roles. They were aware of crossing a boundary from everyday life, and some adapted their contributions to fit within the research space. Although researchers and public collaborators made efforts to create an inclusive climate, obstacles to impact were identified. Conclusions Considering experiential knowledge as a boundary object highlights that this knowledge has a different form to other kinds of knowledge that contribute to research. To enable impact from experiential knowledge, researchers need to create a space where public collaborators experience epistemic justice. Patient and Public Contribution The Peninsula Public Engagement Group (PenPEG) was involved in the planning and conceptualization of the study, including the development of the ethics application and the interview schedules. One member of this group (Richard Fitzgerald) and one from outside the group (Leon Farmer), were full members of the author team and were involved in the data analysis. Leon Farmer has since become a member of PenPEG. Richard Fitzgerald and Leon Farmer were not involved in the three research studies sampled for this study. Sadly Richard Fitzgerald died during the course of this study.


| INTRODUCTION
'Patient and public involvement' (PPI) refers to forms of partnership and collaboration with patients and members of the public in research. We here use the term 'public collaborator' when referring to members of the public or patients who are involved.
PPI arose partly as a response to flawed research. For example, Oliver 1 exposed how assumptions about disability amongst able-bodied researchers skewed survey questions designed to understand the lives of disabled people, rendering the results invalid. Another argument is that research is wasted when the needs of potential users are ignored. 2 Involvement can increase the epistemological resources available to researchers, by pointing out gaps in their knowledge or sharing how health is experienced by patients and carers.
'Experiential knowledge' is gained through living with health and illnesses, and receiving healthcare. 3 'Experiential knowledge arises when these experiences are converted, consciously or unconsciously, into a personal insight'. 4  Public collaborators are asked to contribute to work, which is carried out within the research community. 5 In this community, knowledge is practised within a social hierarchy. As outsiders, public collaborators cross an epistemic boundary between everyday language and technical scientific terminology. Stories are not part of conventional scientific discourse and within it, stories and 'anecdotes' have less value than 'facts' and 'findings'. 6,7 Public collaborators also cross a physical boundary, when research meetings are held at a university.
Experiential knowledge in research might therefore be considered a 'boundary object'. This is 'an analytic concept of those scientific objects which both inhabit several intersecting social worlds … and satisfy the informational requirements of each of them' (p. 393). 8 While 'experiential knowledge' is not a material object, the forms it takes in research settings are similar or identical to the forms it takes in people's personal lives. However, the social meanings of it will differ across the boundary.
Carlile 9 found that knowledge boundaries emerge between different areas of practice 10 because knowledge is localized, embedded and invested. 9 Research knowledge is localized in relation to the problems faced by researchers in a particular study or research practice; it is embedded in research practices because researchers apply methods recognized as reliable within their field; it is invested because researchers have used time and effort gaining knowledge. Achieving impact from PPI can be a challenge since public collaborators are visitors to this study space.
We propose that Fricker's 11 conceptualization of epistemic injustice is relevant to enabling experiential knowledge to be a successful boundary object in health research. Fricker identified two aspects of epistemic injustice. 'Testimonial injustice' refers to 'a wrong done to someone specifically in their capacity as a knower' (p. 1), 11 for example, when public collaborator's contributions are not taken seriously because they are not practitioners or researchers.
'Hermeneutical injustice' is when public collaborators do not have the conceptual tools to interpret their experiences of healthcare or to contribute their experiential knowledge, for example, when professionals use research jargon.
If health research is to benefit from experiential knowledge, public collaborators need testimonial justice. What they say needs to be taken as worthy of influencing the research. 12 Sociologists have had a longstanding interest in people's everyday ideas about health and illness. [13][14][15] These ideas have been explored in the context of public health and healthcare. Recent ethnographies of PPI have led to a deeper understanding of the processes involved. 16,17 In this study, we investigate the contribution of experiential knowledge to research, applying the theory of epistemic injustice.

| METHODS
This study's aim was to increase understanding of how experiential knowledge is used in research settings. Its objectives were to build a working definition of experiential knowledge, use this to analyse data and identify practical applications of our findings to inform PPI practice. We conducted a qualitative study of the involvement of (i) audio transcripts of meetings between researchers and public collaborators and (ii) interviews with public collaborators and researchers.
Ethical approval was granted by the University of Exeter College of Medicine and Health Research Ethics Committee (Nov16/ D/100Δ8).

| Study sample and data collection
This study was based on a convenience sample of three projects (Table 1). These projects were led by researchers who are not authors of this paper. Public collaborators were paid for their time and travel and had a named administration contact for practical help.
The public collaborators were all potential beneficiaries of the interventions being researched. For each project, the first meeting with public collaborators was selected for analysis. The meetings would have happened irrespective of people's consent to audio recording; however, all participants gave written consent.
Six public collaborators were invited and consented to participate in face-to-face individual interviews. We asked how they were (or not) able to contribute their own experiences, and whether they edited their experiences to fit the meeting context. Participants were offered a £10 voucher for taking part. Five researchers were invited to interviews and four consented to face-to-face individual interviews. We explored their perceptions of public collaborators' contributions. Interviews were conducted by authors who had not been involved in the projects.
Interviews lasted about 1 h in a place chosen by the interviewee.
Interviews were transcribed using pseudonyms.

| Data analysis
Data analysis was informed by Green and Thorogood. 18 First, we conducted a preliminary and inductive analysis of the three meeting transcripts to develop a coding framework for experiential knowledge (Table 2). Once the framework was agreed upon, each transcript was coded in NVivo 11 by two authors. One author who had not been present at any of the meetings (N. B.) coded all three meeting transcripts.
Second, we inductively developed codes for the interview transcripts, using a similar process. All interviews were read to inform the framework. Once the coding framework was finalized, each transcript was coded independently by two authors. The coding framework was adjusted after the first round of double coding, and data were recoded with the new framework (Table 3).
In the third phase, we developed coding summaries by examining the data within each code. This allowed us to compare findings across meetings and interviews. All authors were involved in writing the summaries and selecting quotes to illustrate the findings. The quotes below were chosen due to their illustrative power, drawn from all data sources (see Box 1 for reflexivity considerations).

| Findings
Below we elaborate on our data-informed definition of experiential knowledge. We then present the three themes. Equal treatment ensures that public collaborators' contributions are taken seriously; professionalism shows the work that public collaborators do to be taken seriously; the final section illustrates the epistemic impact of experiential knowledge in situations of epistemic justice.

| Experiential knowledge
Our data-based definition states that experiential knowledge includes both own and others' lived experiences, in the forms of opinions, facts and media information. This knowledge is shared through narrative mediums: stories, seeking clarification, answering questions, commenting and references to own roles within the research space.
Below, the public collaborator conveys knowledge of others' experiences through the medium of a story, while connecting this knowledge to the consideration of screening uptake, thereby making meaning of this story for the research setting.
Public collaborator: I remember a lady who used to live near me who was diagnosed, she was in her 50s, and she was diagnosed with some type of bone cancer and at the time they gave her 6 months to live. And she said this isn't going to beat me. I'm not having this.
Twelve years later she passed away …. but it would depend on the individual person. Whether they think 'oh well I've got it -I will just give up now' or whether they are determined they are going to get through it. In the interviews, experiential knowledge was contrasted with technical or academic knowledge and sometimes referred to as  Two public collaborators and one researcher.

Screening
Health Technology Assessment of a screening interven-tion First of three one-off workshops with public collaborators. All workshops followed the same format and purpose, but with different public collaborators.
To identify relevant outcomes for the protocol and discuss the relevance of health inequalities to screening uptake.
University meeting room, tables set up for three small groups. Two researchers from the project team, one researcher observing the meeting, one PPI facilitator, nine members of the public. Both the observing researcher and the PPI facilitator are authors of this paper.
Two were new to involvement, two had limited experience of it, and five had previous experience of being involved in research.
Two public collaborators and one researcher.
Abbreviations: NHS, National Health Service; PPI, patient and public involvement.
T A B L E 2 Experiential knowledge: Coding framework from the meeting transcripts A statement, story, comment or other utterance comes from a public collaborator and includes a reference or alluding to the role that the person has as a public/patient/carer collaborator. For example, by saying 'I don't want to be difficult', 'this isn't a criticism', 'can I just ask', and so forth.
2. Clarification-Public collaborator asks clarifying questions which are:

Clarification supported by own experience
A public collaborator asks a question that refers to, or is clearly anchored in, their own experience. It needs to be explicit that the experience referred to is their own. For example, 'Would that work? Because when I tried it didn't really help me'.

Explicit reference to own lack of knowledge
Absence of knowledge-could be technical or experiential-the public collaborator explicitly refers to own lack of knowledge, for example, I don't know, I'm not sure and so forth.

Clarification supported by one's own understanding of the topic
A public collaborator asks a question that is anchored in their knowledge about the topic. Their topic knowledge must be clearly referred to, for example, 'So if I understand you correctly, clinical effectiveness is found when you compare this new treatment to what's currently delivered by GPs?'.

Clarification with reference to other people's perspectives/lives including popular culture
A public collaborator asks a question on behalf of others, either because they are prompted or because of own initiative. For example, 'we don't seem to have a problem with X-raying vast amounts of people umm exposing them to radiation for breast cancer screening just wondering why we would consider it for this?'. 3.2 Distancing story Shares a story but they are using nouns that generalize the story, for example, words like 'one' instead of 'I', or 'people' instead of 'we'.

Others' story
Shares a story about a friend, family member, or something they witnessed or were told. This can include observed stories about something a public collaborator has seen, for example, someone's treatment by a nurse, and representative stories where the public collaborator speaks on behalf of others and may be relating stories from, for example, a patient group they belong to.

Concern
Shares a concern or raises a question that relates to their own experiences, for example, but not limited to asking for advice or bringing up a problem with an existing service.

Answering a question 4.1 Answering questions
A public collaborator answers question about his/her own behaviour, views or experience directly. This is brief, can include details of experience but is shorter than a story. How public collaborators address obstacles, includes obstacles and descriptions of how to overcome obstacles.
9. How do researchers address obstacles they experience when involving patients and members of the public in research?
How researchers address obstacles, includes obstacles and descriptions of how to overcome obstacles. 10. How do researchers perceive the value of experiential knowledge?
Text contains a value judgement, for example, 'good', or 'perhaps it was a bit narrow'. Use this code also when the researchers talk about no value or lower value of involvement.
11. How do researchers describe public collaborators' role? Text from researchers that clearly relate to how they see the role of public collaborators in research.
12. What does involvement 'good practice' look like, from the perspective of involved patients/public members and researchers?
Text contains a value judgement, words that evaluate the described practice, for example 'I really liked that', or 'in this way they clarified to me'. 13. What does involvement 'poor practice' look like, from the perspective of involved patients/public members and researchers?
Text contains a value judgement, words that evaluate the described practice, for example, 'this was quite difficult because they didn't explain', or 'I wasn't sure what they meant'.
14. (How) does the contribution of experiential knowledge affect the design or conduct of research studies?
Text that speaks about changes in the research made as a result of the patient and public involvement. Our findings suggest that accessible rooms, the layout of furniture, ground rules, reimbursement of expenses and so on, can contribute to public collaborators feeling more equal at meetings.
Good facilitation helps to create a receptive environment, and signals to the public collaborators that their contributions are valued.

| Professionalism
Literature on involvement has brought forth the 'professionalisation paradox'; when public collaborators require training to contribute to research, but this learning compromises their 'lay' status. 22  In relation to hermeneutical injustice, the use of jargon is a frequent obstacle to PPI. Some researchers supplied jargonbusters and lists of acronyms, met with public collaborators in community settings and held premeetings to consider the agenda in advance. Some public collaborators also described doing research at home after the meeting to look up terms that they did not understand.
In some cases, a public collaborator's existing skills and experience from previous employment may inform their contributions. This was seen in the Datasets project meeting:

BOX 1. Reflexivity
Reflexivity context: The authors include four members of the NIHR Applied Research Collaboration South West Peninsula (PenARC) PPI team and one member of the Peninsula Public Engagement Group (PenPEG). Three authors supported involvement at the meetings within the study's data set. They did not conduct interviews with participants from these meetings.
Reflexivity considerations: • The author team could be insufficiently critical of own or colleagues' practices. • Public collaborators may be unwilling to criticize researchers' work.
Reflexivity activities: • At study inception we conducted a team reflexivity exercise. 19 • All authors wrote a reflective statement about their interest in the topic, how they expected to influence the study, their expectations and foreseeable obstacles. • Discussion of statements did not reveal any major disagreements or conflicting aspirations, but emphasized the team's wish to improve their own practice LIABO ET AL. Interviewer: You mean why people accept or don't accept?
Researcher: Exactly. And this allowed us, the PPI allowed us to talk about that, and say, 'We did this exercise and this gave us information about participation and nonparticipation'.
Interviewer: A hugely important issue for a screening programme.
Researcher: Absolutely, absolutely, and, of course, in many cases you don't find about that until you've introduced the screening programme. collaborators asked questions. It may be that any query or alternative perspective feels challenging to some researchers. Public collaborators are aware that challenges need to be made carefully, and in the interviews, they may have claimed to have been more challenging than they were in practice.

| DISCUSSION
The lower value attached to experiential knowledge and personal stories in scientific settings, because they have crossed the boundary from everyday life, creates the potential for epistemic injustice. This study has shown what experiential knowledge contains in terms of personal knowledge of own and others' experiences. We have also shown the mediums that this knowledge is shared through. Our study suggests PPI impact relies on meetings that provide a context of epistemic justice, in which public collaborators' experiential knowledge is valued.
Experiential knowledge has been defined as 'pragmatic rather than theoretical or scientific, … oriented to the here-and-now-action … and holistic' (p. 449). 3 Experiential knowledge spans embodied knowledge gained through personal lived experiences, and empathetic knowledge gained through long and intimate associations with people cared for. 25 A related term is 'lay knowledge', defined as the articulation of meanings people ascribe to health, illness, disability and risk. 13,14 Our study adds to these works by emphasizing what it contains and how it is shared in the context of health research.
Popay et al. 13  Knowledge is informed by the social practices we engage in as humans, embedded in complex social and physical structures.
Epistemic justice relates to our capacity as a giver of this knowledge, a capacity which is intrinsic to human value. 11,29 'Experiential knowledge', as observed and described in three research meetings, speaks to this intersection of forms and sources of personal experience that together become knowledge. 'Experiential knowledge' brings to research public collaborators' lived, felt and embodied knowledge as practice.
In this study, public collaborators were conscious that in researcher-led meetings they had crossed a boundary, and some adapted their contributions accordingly. The mediums that facilitated boundary-crossing were stories, comments and questions. For example, we observed in our meeting data and researchers told us in interviews, that naïve questions can trigger a deeper understanding by researchers of the work they do. These questions can be asked on either side of the scientific boundary, but their function in the research meeting is different than in personal or social situations.
As highlighted by Fox, 30 boundary objects may have positive or negative social meanings and clearly, patients' stories have the potential to be seen as disruptive of scientific discourse. A successful boundary object needs to have a shared syntax to represent knowledge and be a means for individuals to specify and learn about differences and dependencies across a given boundary. 9 We propose that in PPI, experiential knowledge can be a successful boundary object that enhances researchers' understanding of their field and how to research it. Our data also suggest that the extent to which experiential knowledge is a successful boundary object depends on whether public collaborators experience epistemic justice.
In research settings, the personal experiences of public collabo-

| STRENGTHS AND LIMITATIONS
The strengths of this study lie in the complementary sources of data from three contrasting projects and the active involvement of public collaborators all the way through from the original proposal to writing this paper. The limitations are that the team researched its own practices, and the study was based within a single institutional context. The involvement support was similar across the projects but the projects themselves were very different and led by different researchers. A similar study conducted elsewhere might have surfaced different concerns, for example, if the public collaborators did not have one named contact for practical queries, or if the researchers at the meetings were less committed to involvement.
The specific institutional context is also a strength. It shows that in spite of considerable efforts to support public contributors to attend and participate in meetings there were still barriers to experiential knowledge being shared and taken seriously.

| CONCLUSION
This study aimed to increase understanding of how experiential knowledge is used in research settings. It found that public collaborators brought experiential knowledge in the form of their own personal experiences, experiences of others and knowledge gained through occupation or media. Public collaborators were mindful of their lack of research knowledge, and this awareness was brought to the meetings, as well as their knowledge that a patient's expertise is usually granted lower status than a professor's.