Imposter Participants? Towards a Reflexive Epistemology of ‘Suspected Participants’

Abstract

There is an increased reliance, since COVID-19, on online methods in qualitative research for participant recruitment and data collection. While online methods have some clear advantages, they have also raised ‘new’ concerns about fraudulent participation in qualitative research that can hamper research rigour. Drawing on a case study of two of our own UK-based projects which investigated intersections of health, sexuality, and social connections during and after the COVID-19 pandemic, this paper critically examines the challenges posed by terms like “imposter” and “fraudulent” participants and suggests instead the use of “suspected participants” to acknowledge the researchers’ role in judging if participants are authentic or not. We argue for a reflexive epistemology that questions normative assumptions about authentic participation and advocate an inclusive but separate analysis of suspect data as an interpretive tool that can enhance rigour, rather than rejecting such data as a problem that needs ‘screening out’.

Introduction

The phenomenon of ‘imposter’ participants, who provide false or inaccurate information in order to take part in research, is thought to pose a significant challenge for researchers, especially in health research, which seeks to inform policy and practice. In this paper, we consider this challenge in relation to online qualitative research which has become more common since COVID-19, and which has led some researchers to raise concerns about fraudulent participants that hamper research rigour. Taking a case study approach, the paper reflects on our experience of two projects that used qualitative online interviews to explore the intersections of sexuality and health during and after the COVID-19 crisis years in the UK. The first project, entitled “Dating App Connections” and funded by the UK Economic and Social Research Council (ESRC), sought to examine how dating app use changed amongst people with different sexual identities over the period before, during and after the COVID-19 lockdowns and social restrictions. It included a national survey of over 830 respondents and 53 online semi-structured interviews, and sought to address if and how the use of dating apps impacted on wellbeing, loneliness and risk-taking. The second project, entitled “Viral Memories” and also funded by the ESRC, explored gay men’s experiences of HIV and COVID-19 by conducting 50 online semi-structured interviews with gay men who lived in the UK. It aimed to identify their specific health and support needs and to recommend strategies to meet these.

This paper critically engages with the phenomenon of ‘imposter’ or ‘fraudulent’ participants in online interviews, proposing instead the term ‘suspected participants’ to foreground the role of researchers deeming something suspicious. We argue that the term ‘imposter participant’ is problematic because it involves the researcher issuing an expert claim by judging (or assuming they know) the ‘truth’ of potential or actual participants’ identities, experiences and motivations for taking part in research. This is ethically problematic as it can adjudicate moral corruption to potential participants while bolstering the researchers’ own claim to ethical rigor. It can also be informed by normative prejudices and exclude some participants on the basis that their stories fail to conform to what researchers believe to be convincing and coherent narratives. Finally, over-wariness of ‘imposter’ or ‘fraudulent’ participants can infuse the relationships between researchers and participants with an unwarranted degree of distrust. Considering these problems, we argue that the emphasis should be shifted away from imposter and fraudulent participants to ‘suspected participants.’ This move more accurately reflects who ‘owns’ the problem of suspicion – the researchers and not the researched. We define suspicion as the researchers’ feeling that a participants’ account of their identity, practice or experience may be fabricated to enable participation in the research, be it for financial gain, to alter the research findings, or for any other reason. ¹

The key question at the heart of this paper is how researchers should approach cases where participants’ accounts of their identities, motivations, or narratives appear questionable. Unlike existing literature that often frames suspected participants as threats to data integrity, this paper argues for a reflexive epistemology that treats suspicion as a productive site of inquiry rather than a methodological problem to be eliminated. Our main argument is that, epistemologically, the notion of imposter participants requires a more reflexively established foundation than is usually provided in methodological discussions of the phenomenon to date.

In terms of structure, we begin by presenting examples of encounters with suspected participants from a case study of own research. We then review existing literature on imposter or fraudulent participants, highlighting its epistemological shortcomings. The subsequent section develops a reflexively deconstructive and reconstructive epistemological framework. Based on this, we suggest that researchers should consider and critically engage with suspected participants by including them in the research, rather than excluding them outright. One way that this can be achieved is by subjecting the suspect and assumed-to-be authentic data to separate analyses and putting them into conversation or comparing them in the overall analysis of the research. Overall, we argue for the need for qualitative online (and other) research communities to engage in greater depth with reflexive epistemology (cf. Hamati-Ataya, 2014) to more adequately grasp and address the problem. The paper concludes by outlining practical and theoretical implications of this approach for qualitative research communities. Rather than excluding suspect participants and the data they generate outright, the approach we suggest can improve rather than diminish rigour. We acknowledge that there are costs to such an approach in terms of the researcher’s time and finances, but believe the advantages in terms of rigour can outweigh these.

Our Experience With Suspected Participants

In the Dating App Connections project, survey data collection was contracted to a reputable third-party provider (cf. Garcia-Iglesias et al., 2024). Survey respondents were then invited to take part in an online qualitative interview. Following this, and to extend and diversify our interview sample, community organisations across the UK, specialist media channels, and institutional social media profiles were used to recruit additional interviewees. Interview participants were informed in advance that they eligible for a £20 shopping voucher as ‘thank-you’ for their contribution to the project. The Viral Memories project relied initially on community organisations for recruitment: these organisations used their internal channels (mailing lists and the like) to share recruitment materials with potential participants. Like the Dating App Connections project, participants were given a £20 ‘thank-you’ shopping voucher after interview completion. Subsequently, social media channels were used but did not include mention to any ‘thank-you’ rewards for participation. In both projects, interested participants completed an online registration form before being contacted to schedule an interview.

Like Santinele Martino et al. (2024), in both projects the teams began to receive an influx of emails from potential participants when the advertisements went public: as many as ten or fifteen in a few hours. Many of these emails followed a similar structure and had the same or similar subject lines. They used similar language and were often sent from similar generic email addresses (such as surname-plus-twodigits-at-gmail.com or hotmail.com). If not replied to within a few hours, follow-up emails often arrived. One such email read:

Hello, I am writing to indicate my interest to take part in the study. My experiences are in line with this research project and I am also inclined to share them as my contribution to improving this subject area. I would love to understand the research better and how it would involve me. Do provide more details about participation. I look forward to hearing from you. Thank you.

At first glance, there was little reason to suspect these expressions of interest, and we initially took them at face value (in fact, we were delighted at such interest in our projects). However, as we moved forward with scheduling and conducting the interviews, issues emerged that raised concerns about the nature of some of these participants’ motivations for taking part in the research, their identities, and the experiences they recounted. During interviews, some participants we already had concerns about provided only vague answers to our questions. For example, some were unable to give the name of any dating app they had used, or say when and how they used it, or describe their experiences of it. At times, participants’ responses would be inconsistent within the same interview, saying that they were both avid users of dating apps and had never used them, or that they were both living with HIV and had never been tested for it.

Discrepancies also arose in some online identities: for example, names provided in registration forms did not match the names provided in the interview or in their Zoom profiles. Some participants stated they had seen the recruitment call on Facebook (despite neither the teams nor its partners using this platform for recruitment). On two occasions, the same participant, under different names, sought to take part in two different interviews. Taken in isolation the researchers could have satisfactorily explained many of these issues to themselves. However, taken together they raised our concerns about the motivations and experiences that some participants reported and the implications for the quality of the data.

In reflecting on these issues, we began to wonder: Who were these participants whose identities, motivations, and experiences we have come to suspect? Was it their intention to deliberately mislead us, and if so, was this primarily linked to monetary gain? Was it morally right for us, as researchers, to subject potential/actual participants’ narratives of their identities, experiences and motivations to interrogation, and to judge if they should be excluded from the research? In addition, and critically, the temptation amongst those members of the teams in direct contact with participants was to approach interviewees with suspicion (and even a degree of frustration), which could introduce an unwanted dynamic in the research relationship.

These questions and issues were exacerbated by the very limited literature on the subject as it concerned qualitative online research: both teams struggled to find substantially detailed accounts of similar issues in published discussions of qualitative online methods (but see Lawrence et al., 2023; Ridge et al., 2023; Roehl & Harland, 2022). The majority of published work was in relation to quantitative survey work, where it seems the term ‘imposter participants’ emerged from, which did not speak to qualitative approaches to social sciences and health research, which are increasingly carried out online (but see Ridge et al., 2023).

Constructing Imposter and Fraudulent Participants

In a review of the existing literature, we noted that where the notion of ‘imposter’ participants is used, it is rarely explicitly defined. In most cases it is used to refer to participants who misrepresent their identity and/or experiences to take part in research, potentially for monetary compensation. Discussions of imposter participants in quantitative research span a wide array of disciplines, particularly health (geriatrics, infectious disease, paediatrics) and behavioural sciences (psychology, psychiatry), but also other social sciences (Owens, 2022). While the literature has focused mainly on quantitative methods, particularly online surveys and clinical trials, the recognition of “the unique challenges and experiences of facing [imposter participants] in the context of qualitative research” (Santinele Martino et al., 2024, p. 1293) is just beginning to come to the fore.

Although some isolated discussions of imposter participants in qualitative research predate 2020, most articles on imposters in qualitative research seem to have been published in the wake of the COVID-19 pandemic, suggesting an increase in, or increased sensitivity to, imposter participants since 2020. Notably, discussions of imposter participants prior to the COVID-19 pandemic did not define the issue as a ‘phenomenon’, as it seemed to be less frequently noticed or acknowledged and less efforts were made to explain it. Some authors have argued that the pandemic’s impact on economic stability (and long-term inflation and ‘cost of living’ crises), as well as the increased reliance on online methods for recruitment and data collection—which reached unprecedented use during COVID-19 era––are key factors in the increase in imposter participants (Lawrence et al., 2023; Ridge et al., 2023; Roehl & Harland, 2022). Newman et al. (2021) suggest that it is possible that the pandemic created a situation where marginalized and vulnerable individuals faced economic circumstances so critical that they were willing to participate in a research study solely for compensation. It is unlikely that financial need combined with online methods are the only explanations for imposter participation: other motivations, such fantasy, the desire to occupy an imagined persona, boredom, or a genuine desire to disrupt research for personal or ideological reasons could also be relevant.

In discussions of quantitative research, there is a wide range of terminology used that chime with the notion of ‘imposter participants’ (Ridge et al., 2023; Roehl & Harland, 2022; Santinele Martino et al., 2024), including “fraudulent participants” (Davies et al., 2024; Lawrence et al., 2023; Sefcik et al., 2023), or “fraudulent survey takers” (Salinas, 2023). Other authors, instead, focus on the acts, using terms such as “participant dishonesty” (Hydock, 2018) or “suspicious participation” (Lawlor et al., 2021). Finally, the data produced are also sometimes referred as “fraudulent entries” (Guest et al., 2021), or “suspicious data” (Salinas, 2023). Thus, the language surrounding imposter participants often carries adversarial and moral connotations, reflecting the perceived threat they pose to research integrity. Descriptions such as imposter participants “attack[ing]” research (Davies et al., 2024, p. 1313) or the above listed “dishonesty” (Hydock, 2018) reinforce this. We argue that instead of these terms ‘suspected participants’ is more adequate as it places the suspicions more accurately at the foot of the researchers and not the potential or actual participants.

Much of the existing literature on suspected participants seeks to identify “red flags” (indicators that raise the researchers’ suspicion about a participant) and proposes solutions to prevent or mitigate their impact on the research. Santinele Martino et al. (2024) and Ridge et al. (2023) outline several indicators that qualitative online researchers should be aware of, such as blank subject lines on emails, non-personalised messages, or an unwillingness to turn their camera on during interviews. Lawlor et al. (2021) go further in developing what they term the “REAL framework” for survey research which refers to how researchers should reflect, expect, analyse, and label “fraudulent” participants “to plan for addressing survey fraud and making determinations about the inclusion or exclusion of participant submissions from the dataset based on level of suspicion” (1).

A multitude of solutions, often technically complex, are proposed, including recording IP addresses to prevent multiple entries from the same address; using cookies to identify individuals attempting to complete a survey more than once; requiring that participants turn their camera on during interviews; or asking participants to provide a working phone number (Hydock, 2018; Jones et al., 2021; Lawlor et al., 2021; Ridge et al., 2023). Such solutions are often intrusive measures that can negatively impact participants’ experiences and trust in the research, and hinder rapport building between participants and researchers. Furthermore, some of these are unlikely to be acceptable to ethics committees on the basis of collecting ‘unnecessary personal data’, and to participants when researching stigmatized identities or practices.

In research related to sex, family, or sensitive topics or identities, participants may be unwilling to turn on their camera or provide their phone number for reasons linked to fear of harassment and to protect privacy, not because they are ‘imposters’. This is something that we experienced in our own projects when interviewing people who felt stigmatised because they were using dating apps or identified as gay men living with HIV. For example, in the Viral Memories project a participant recounted that he wanted to: “keep my identity private like nobody recognizing my face in anything because I am [HIV] positive and then I’m owning up to this. So, it’s a very private and sensitive topic”. The application of many of the previously proposed ‘solutions’ to imposter participants could, in fact, exclude the voices of those groups who are already underrepresented in research, such as people who struggles with loneliness and wellbeing, whose sexuality or health status is ‘hidden’ for self-protection, or who experience shame or stigma.

All this illustrates the difficult balancing act that researchers must undertake when dealing with suspected participants between inclusion and safeguarding the rigour of research data. On the one hand, researchers need to safeguard participants’ right to privacy and aim to maximize inclusion by removing barriers to research. As Martino et al. (2024) ask: “are we excluding eligible participants due to our suspicion and creating barriers for participants who are so often questioned and silenced?” (1295), and, as Jones et al. (2021) note, it is important not to discard unusual but legitimate responses.

Excluding eligible participants can inadvertently silence voices that don’t ‘fit’ normative expectations and that are already marginalized. For example, in our work on dating apps, participants whom we suspected of being ‘imposters’ often presented themselves as Black African, living with HIV, and trans people. These were groups we had specifically targeted for recruitment as they were underrepresented in the study. Our suspicions were sometimes raised by their refusal to turn their camera on during an interview or provide a phone number. However, such refusals may well have been related to the stigma and risks of harassment or violence that they feared, and not to them being imposter respondents.

This raises the ethics involved in how researchers make decisions about the inclusion and exclusion of potential participants: through our reflections on our own ‘impulses’ to exclude suspected participants, we recognise that erecting barriers to prevent less visibly verifiable or apparently coherent participants may also exclude legitimate participants, silencing their voices and diminishing the richness of a study. The question that arises for us in terms of future research is: how can this be reconciled with the ethical imperative to avoid including ‘imposters’, given that the “the inclusion of fraudulent responses diminishes the value of legitimate responses”? (Lawlor et al., 2021, p. 7). The balance between inclusivity and rigour is a delicate and complex one, and the current literature on imposter participants fails to provide an adequate framework to critically address this phenomenon. We believe there is a need for a greater engagement with reflexive epistemology with respect to suspected participants and how we respond, as researchers, to our own concerns.

Approaching Suspected Participants Epistemologically

In this section we reflect on how we sought to develop an approach to suspected participants based on an interpretative-constructivist epistemology and highlight its limitations. Epistemology, as the theory of knowledge concerning methods, validity, rigour, and scope underpins the methodologies and theoretical frameworks we deploy to answer the questions we ask. Thus, the epistemologies of qualitative social science refer to the frames and lenses used to analyse and interpret the data. In relation to this, Flicker (2004) argues that the validity of research is not absolute, but rather depends on “how we imagine our role as researchers and the possibilities for our research findings” (534) This, in turn, will generate “different approaches” to managing and interpreting data (ibid.).

In this regard, qualitative social science researchers are focused on “describing, interpreting and understanding the meanings which people attribute to their existence and to their world” (Cutcliffe & McKenna, 1999, p. 375). While doing so requires developing epistemologies that are concerned not with a positivistic search for ‘the truth’ but with trustworthiness and authenticity of data (Lincoln & Guba, 1985), at the same time it is important to acknowledge that lived experience can be messy, contradictory, and blurry (Heaphy, 2018; Pope & Mays, 1995). This has long been accepted in qualitative social sciences, including through the so-called “narrative turn” and interpretative-constructive approaches. For example, Heaphy et al. (1998) argue that interviewees’ responses represent “a particular narrative […] that is likely, at least to some extent, to be framed in the language that the researcher brings to the interview” (461). The constructed nature of data is also emphasized by others, such as Goffman (1959) or Lune and Berg (2017). Holstein and Gubrium (2012) for example, explain how interview participants may “tell lies or just set out to please the interviewer” or may struggle to recall past events, instead reconstructing them “in order to make sense of their present selves” (99).

Much of the existing literature on imposter participants, which arises from behavioural or health sciences, often implies that imposters are ‘lying their way’ into research and compromising data. It seeks to develop new tools, both technological and methodological, to exclude them in the hopes of obtaining ‘truer’ data. As qualitative social scientists, to adopt an approach that sees imposters solely as fraudulent or as ‘liars’ that should be screened out means embracing a positivistic epistemology that is at odds with our disciplinary goal of embracing the interpretative and constructed nature of data.

An alternative approach, more aligned with the epistemology of qualitative social sciences, may be found in Garcia-Iglesias’ (2022) study of ‘bugchasing’ - the eroticising of HIV infection. Garcia-Iglesias interviewed a variety of gay men who purportedly eroticised and sought out HIV infection. In reflecting on the interview data, he noted that some of his research participants may have exaggerated their accounts or may have volunteered to be interviewed as a way of satisfying their sexual desires: narrating stories that were arousing to them but may not necessarily have been an accurate account of their experiences or practices. Nonetheless, Garcia-Iglesias does not label them as ‘imposters’. He suggested that “rather than think of [them] as lying, it could be said that [they] may be evidencing just how much fantasy and reality are intertwined” (48). Like Heaphy et al. (1998), Garcia-Iglesias draws on Plummer’s approach to data as personal stories and the goal of exploring “the ways in which they are produced, the ways that they are read, the work they perform in the wider social order, how they change, and their role in the political process” (Plummer, 1995, p. 19). From this perspective, the concern was less about whether the personal narratives provided by participants were an accurate depiction of their lives and actual practices than about the insight that those stories could provide into the social, cultural and political meanings of bugchasing.

Nevertheless, “even those researchers steeped in interpretative traditions remain concerned about the trustworthiness and authenticity of their data” (Lincoln & Guba, 1985 cited in Flicker, 2004, p. 533). That is, an interpretative epistemology does not mean a ‘free-for-all’: there remains a profound commitment to rigour, oftentimes through critical explorations of the data and its constructed nature. In fact, in our own research context we felt that there was somehow a difference between the suspected exaggeration or misrepresentation we had encountered in our previous work and the kinds of practices associated with suspected participants. As much as we saw ourselves as deeply embedded in an interpretative-constructivist epistemology, we struggled to feel that suspected participants were providing useful data that was not damaging to our research. We felt that exaggeration, for example, might be a finding in itself, whereas imposter participants stretched the boundaries of ‘interpretivism’ too far.

We believed that having no framework to understand and navigate imposter participants could lead, as it did in our case, to questioning genuine participants and approaching all participant encounters with suspicion, preventing rapport. In addition, we also felt the need, as stated above, to develop strategies that balanced participants’ right to privacy, in order to ensure that we are not excluding ‘legitimate’ participants, yet that also safeguard our methodological rigour and data. Reflecting on this, we propose a multi-layered approach to suspected participants in online qualitative research that is grounded in reflexive epistemology.

Reflexive Deconstructive and Reconstructive Epistemologies of Imposter Participants

In this section we provide a background for reflexivity and reflexive epistemologies and suggest how a reflexive approach to imposter participants might look like. The approach we advocate for has emerged from considering three overlapping elements of reflexivity: reflexive method, reflexive theory, and reflexive epistemology. At its most simple reflexive method in qualitative work commonly refers to the efforts that researchers make to reflect on and make explicit how research is designed and undertaken with rigour in mind, as opposed to a positivistic notion of objective validity. It often involves discussions of the socio-cultural and professional positions of the researchers, as well as reflection on how various research relationships shape the research results (Heaphy, 2007). In terms of analysis and presentation, the emphasis is often on interpretation, and the frameworks that have guided this. This raises the issue of theory.

Reflexive theory addresses how researchers adopt and integrate theoretical frameworks (Heaphy, 2007). Some employ a singular theoretical lens, while others draw from a ‘toolbox’ of concepts. In both cases, what matters in reflexive approaches, is that researchers are explicit about their theoretical influences and how this has shaped the methods they chose and their interpretation of the data. However, there is also another approach to reflexivity in theory, one that might be better termed ‘theory of reflexivity’ (ibid). This focuses on how reflexivity is also an aspect of everyday life, and how research interpretations should be acknowledged as ‘interpretations of interpretations’. This approach has developed partly in response to radically deconstructive theoretical approaches that question all claims to ‘truth’ or ‘reality’ and that view knowledge through the lens of dynamics of power where those who have access to social, cultural and financial resources (in the case of research, the researchers) can ultimately claim authority over the authenticity and truth of the identities, experiences and practice that they research.

Reflexive-deconstructive epistemologies go further in viewing epistemology itself as historically and socially constructed. More significantly, while they entail deconstructing hegemonic epistemologies, they are also involved in a process of reconstruction (Heaphy, 2007). Hamati-Ataya (2014) articulates reflexivity as both a “bending back” (critically assessing knowledge within its socio-historical context) and a “bending forward” (using this understanding to enable meaningful and responsible scholarly practice). Similarly, Bourdieu (1997) argues that sociology, like all established orders, naturalizes its own arbitrariness, creating a ‘sense of reality’ that research communities must actively and collectively challenge through reflexive engagement. Considering the above discussion, we now turn to implications for suspected participants.

A Reflexive Deconstructive and Reconstructive Epistemology of Imposter Participants

In the case of suspected participants, reflexive method can involve explicit reflection on how the researchers’ positioning, assumptions and expectations, have implications for who is deemed a ‘legitimate’ participant, and what are judged as feasible or coherent narratives of identity, experiences and practices. It could also involve reflection on the impact of financially rewarding participation in research or other material benefits. More broadly, reflexive method would involve an openness on the part of the researcher about the problems encountered and the efforts made to manage and address these. Such discussions aim for transparency as a way of establishing rigour. In this case, such discussion serves to illuminate and provide suggestions for identifying, managing and minimising the risks and challenges that suspected participants present to rigorous qualitive research. While such approaches can be valuable to those who see ‘imposters’ as a problem that requires management and exclusion, they are methodologically limited by the extent to which they leave the theoretical and epistemological assumptions that underlie the concepts of imposter, fraudulent, or suspected participants themselves unchallenged. Thus, we argue for the need to also consider reconstructive reflexive approaches in relation to suspected participants.

Deconstructive reflexive theoretical approaches go much further in challenging the assumptions that underpin the reflexive method approach. Such approaches draw on theoretical arguments about how identities, biographies, memories, motivations and frames of everyday understandings are radically open, if not fragmented. They question the very notion of authentic selves and experiences, and acknowledge the existence of incoherent, contradictory and performative (narratives of) selves. From this perspective, suspected participants, whether they are conceived as being self-aware actors or not, can be viewed not as imposters, but as expressing the open, contradictory, performed and performative nature of identity and experience that is a feature of post-modern societies. In line with these approaches, efforts should not be made to exclude them in the name of rigour, but to develop research methods anchored in deconstructive epistemologies, which are better attuned to post-structural or reflexive theories of experience. The implications of this are a critical questioning (or deconstruction) of the assumptions that underpin our theories of knowledge and the practices that flow from them: how as researchers and research communities we judge research as rigorous and convincing and evaluate research accounts for their place in the flow of power. It is only by engaging in such exercises that we become involved in a legitimate reconstructive sociology.

What might be the implications of suspected participants for sociological praxis? One answer, from a reconstructive methodology and reflexive epistemology perspective, may be counter-intuitive: not to exclude suspected participants from research but rather to include them. This does not imply ignoring researchers’ concerns about specific participants or their accounts. It involves, instead, explicitly addressing these: we suggest labelling and separating them out as accounts that are subject to the researcher’s suspicion. At the same time, and in turn, we suggest remaining open to the insights that such suspected accounts might provide for the research, perhaps even for questions that the researchers did not explicitly set out to answer. By including but separating out this data, we may also be opening the door to interrogating more fully our claims to knowledge and—in turn—to the flow of power that underlies our research practice: this includes taking responsibility for our decisions about what identities, narratives, findings and methods we deem as acceptable or ‘authentic’; the factors that shape our responses to participants; the power dynamics we are involved in enabling and reproducing knowledge. In doing so, we may realise the limits of inherited and hegemonic epistemological frames and begin to bend forward towards alternative interpretations of social reality, and to develop convincing sociological praxis.

It could be asked if this goes against the ethic of rigour that is central to ‘authority’ of qualitative research? Quite the opposite: it may enhance rigour. Reflecting on the case study of our own research, we can see how and where the analysis of the data generated by those we deemed authentic and those deemed suspicious would produce similar and different research findings, and if and where we might need to revisit our initial analyses to clarify our interpretations of the ‘authentic’ data. We can also see how suspected participants might imagine the experience of dating apps for the subjectivities, experiences and practices we were interested in, as well as social conditions, risks and supports that are available to ‘othered’ groups (such as gay men living with HIV, and marginalised sexualities, and subordinate genders). This could allow insights into how such groups are ‘othered’ in diverse and broader cultures, and the power dynamics that the ‘othered’ making audible their authentic experience. On the basis of our own case studies, in the following conclusion we suggest a number of points that online qualitative researchers would do well to take into account in guiding their response to participants they view as suspicious.

Conclusion: A Reflexive Epistemology of Suspected Participants

In this conclusion, we attend to the practical import of a reflexive deconstructive and reconstructive epistemology of suspected participants. In doing so, we start from the position that suspected participants are not simply a threat to research rigour, and that practical responses to suspected participants are not limited to identifying ‘imposters’ or ‘frauds’ and excluding then. We should be cautious about viewing and constructing suspected participants only as a threat to quality research and rigour. For us this is problematic ethically and in the claims to authority over judgements of authenticity that such constructions imply. It is also problematic, as the data they generate could also be used to enhance rigour. The predominant concern amongst researchers engaged in online qualitative research with developing practical responses to suspected participants is important. However, so far, this has been insufficiently grounded epistemologically, an issue we have aimed to address here. Building on this we suggest the value of online qualitative researchers’ awareness of the following:

Suspected Participant Data can Be Relevant to Research

Suspected participants should not to be banished as useless or harmful to the research, but rather we suggest that they could be considered as a distinctive, yet relevant part of the data generated. These data could provide a unique opportunity to generate insights into the reflexive or fragmented nature of identities and associated and the experiences and practices that shape them; survival strategies in the face of economic, social and cultural exclusions and inequalities; and attempts at recognition and social participation. They can illuminate imagined ‘realities’ and the links and disjuncture between cultural imaginations and lived realities. Suspected participants need not to be included as ‘authentic’ narrators as envisaged by researchers but may well be narrators of other authentic experiences and realities that provide an alternative view of the research itself.

Suspected Participant Data can Serve as Comparison Points

The data generated by suspected participants can also provide a comparative point between ‘trusted’ accounts of experience and those deemed to be more dubious. In this respect, we might compare ‘actual’ and ‘imagined’ experience, ‘reality’ and ‘fantasy’, as well as the relative value of the rewards offered to participants: be they financial, material, cultural recognition, pleasure, political representation and so on. In this way, the balance of the boundaries between inclusion and exclusion can be maintained: with the voices deemed as authentic forming the (contingent) reality of experience, and the voices deemed to be suspect as a comparative point against which we judge the success of our methods, theoretical and epistemological frames that give rise to our analysis.

There are Challenges as Well as Possibilities in ‘Including’ Suspected Participants in Research

Of course, resources could present an important obstacle to truly reflexively epistemological approach and the ensuing analyses in terms of time, funding, financial costs. This would be especially so if researchers were to take on the burden of comparative analysis of seemingly authentic and suspect voices. Research happens in a context of underfunding, short projects, regulatory requirements such as ethics approval, and casualization, that makes it difficult to find the bandwidth for comprehensive engagement with reflexive epistemology. While there is already a toll that is endured by researchers in participant-facing roles (often the most junior team members) the burden of this work is likely to fall on those who lead the design and analysis of research. Also, while there may be cases where such comparative analysis may be suitable for publication, this will by no means always be the case.

Nevertheless, the possibility of expanding our vision of empirical realities is, in itself, worthy of exploration. Where this is not feasible, there is a need at the very least to acknowledge that our research is limited by hegemonic epistemologies, and the problems, as outlined in this paper, of constructing suspected participants in derogatory ways as ‘imposters’, ‘fraudulent’ or ‘liars’. It is incumbent on the interpretivist social sciences, and particularly in online qualitative research —as we have aimed to show in this paper—to inform the agenda on suspected participants as opposed to imposing other more positivist inclined disciplines’ epistemologies upon ourselves.

These recommendations are examples drawn from our own case studies, but we acknowledge that many other alternatives are possible. We suggest that each researcher or team, in conversation with the epistemological literature, theory and other researchers who are engaged in online qualitative research, should seek to reflexively deconstruct and reconstruct their epistemologies in ways are relevant to own concerns about suspected participants.