I'm Okay,You're Okay?: Reflections on the Well-Being and Ethical Requirements of Researchers and Research Participants in Conducting Qualitative Fieldwork Interviews

Introduction

In this paper we present our reflections as two researchers who worked together on a U.K. government–funded study that explored the employment experiences of people who had claimed incapacity benefits because of mental ill health. Conducting research on a sensitive theme with a potentially vulnerable group of people gave a renewed focus to some perennial issues for social research, including consent and control, rapport building, managing and responding to emotion, and offering appropriate longer term support. The project highlighted that these issues have relevance for both participant and researcher well-being and also have a number of ethical implications for both parties.

We will explore here some of the practical, methodological, and ethical issues that were raised while planning and conducting qualitative interviews in the field and discuss how these were addressed, if not always resolved. We will also highlight some potential challenges faced by contract researchers, whose work is frequently driven by external interests and shaped by limited time and resources.

Following a brief introduction to the study that prompted these issues, we will present our personal reflections, practical steps or procedures found to be helpful at each stage of the fieldwork process, and the questions that remained unresolved. These are discussed under four thematic headings:

We will conclude the paper by emphasizing the continuing need for researcher reflexivity and flexibility in conducting social research and raise some questions and suggestions that other researchers might like to reflect on as they strive to balance the needs and demands of participants and ethically appropriate research with their own well-being as professional researchers.

Background to the study

The project on which this paper draws was a study of the employment experiences of people who had claimed an incapacity benefit for mental health reasons (Sainsbury et al., 2008). Mental health is currently high profile and much discussed in U.K. public policy, with concern surrounding the emotional, social, and economic costs of mental ill health to individuals and their families. These concerns include failures to meet the employment aspirations and support needs of people with mental health conditions (Healthcare Commission, 2007; Seebohn & Secker, 2005), the widespread prevalence of “common” mental health problems such as depression and anxiety (Centre for Economic Performance, Mental Health Policy Group, 2006), and the economic costs of associated levels of sickness absence, long-term unemployment, and high numbers of people claiming incapacity benefits due to a mental health condition (Department for Work and Pensions [DWP], 2002; Layard, 2004; Sainsbury Centre for Mental Health, 2007). This latter issue provided the impetus for the present study, which focused on the employment experiences of people who had moved from work to claiming an incapacity benefit for reasons of mental ill health, and of people who had made the move from incapacity benefit back into work.

Although the participants' health circumstances gave rise to some specific interview dynamics, we do not feel that the issues discussed below are in any way exclusive to people who have experienced mental ill health. Some of the issues might be salient to people experiencing ill health or mental distress; however, the reflections presented here must be viewed in the broader context of qualitative research encounters. The research draws on mental health as an illustrative rather than problematic example, as many media and social stereotypes tend to depict mental health issues (Pinfold & Thornicroft, 2006; Salter & Byrne, 2000; Seale, 2003).

Negotiating consent

The issue of informed consent remains controversial, with debates highlighting differing perspectives and expectations of researchers and participants as to manner, timing, and depth of information required or desired in obtaining consent (Birch, Miller, Mauthner, & Jessop, 2002; Graham, Grewal, & Lewis, 2006; Lee & Renzetti, 1993). In the mental health and employment project, an opt-out procedure was used in recruiting study participants, whereby potential participants were selected from benefits records, written to, and asked to respond proactively (by mail, phone, or e-mail) if they did not wish to be contacted by a researcher regarding participation in the study. People who did not opt out at this early stage were, of course, able to decline to take part at the point when subsequent contact was made. For one of us this process did not raise any ethical concerns as it was the usual and accepted procedure for projects on which she had previously worked. However, for the other this process of opting out of rather than into research participation was unfamiliar and initially posed some concerns with regard to informed consent. Although recognizing that this process was a practical means of obtaining a research sample from a potentially hard-to-reach group, she felt that the opt-out approach brought to the fore a number of general and more specific ethical issues.

First, at a general level, there was the issue of understanding. Would a formally worded letter instructing people to opt out if they did not want to be contacted be understood by everyone, especially people with low literacy or whose first language was not English? Second, there were specific concerns about how potentially vulnerable individuals, in this case people with mental health conditions, might feel about opting out: Would they feel able to? We were conscious of the dominant role of the sponsoring government department as gatekeeper in the recruitment process. In social research the importance of gatekeepers' power in both selecting potential participants and facilitating their ability to say no is well known (Miller & Bell, 2002). Although in this case the sample was randomly drawn and the identities of participating individuals were not made known to the project sponsor, one of us had concerns that participants might have felt compelled to take part because of their current and/or past relationship to this particular gatekeeper and associated issues of financial vulnerability. Although both written and verbal information about the study stressed that participation (or not) would have no effect on people's entitlement to benefits, this might be a difficult guarantee to accept for people whose financial well-being depends entirely on the decisions of government departments, potentially engendering feelings of obligation or limited choice. Thus, for one of us the concept of volunteering consent was initially called into question by the recruitment approach used in this project as she found herself following a process that was different from previously used procedures of opting in.

However, as the study progressed, the researcher become more comfortable with this process as it became increasingly apparent that research participants were frequently not powerless in negotiating consent. In fact, a small number of people proactively opted into the study, using the contact details supplied for the purposes of opting out to contact the research unit and volunteer to take part. Aside from these volunteers, we both noted a range of strategies, direct and indirect, through which people asserted their wish to opt out or later sought to exercise control in the research encounter (see also Thapar-Björkert & Henry's 2004 personal account). Direct strategies to decline overall participation were to opt out by response to the initial approach letter or later, when telephone contact was made. Indirectly, people who had confirmed or tentatively agreed to an interview employed strategies of not responding to researchers' later phone calls and messages, canceling or postponing appointments (sometimes repeatedly), or simply not being present at agreed meeting times and places. It was also reassuring to observe some participants negotiating consent on an ongoing basis throughout the interviews; for example, choosing to exercise their right not to answer certain questions or making it clear when they had shared as much information as they wished to on a particular theme.

These observations were of comfort to both researchers in demonstrating that study participants saw their right to opt out as genuine and felt able to exercise selectivity and control in what they shared during interviews. Despite this, it is important to remember, as Graham et al. (2006) have noted, that “withholding information during a qualitative research interview may be more difficult to operationalise than participants anticipate” (p. 40). Thus, although there were comments from some participants that the interview encounter had been positive in enabling them to talk through their feelings, we do not know for certain that participants continued to feel comfortable with the extent to which they had shared personal circumstances and experiences with us after the event.

Furthermore, our differing initial reactions raise a wider issue worthy of future consideration, namely, how familiarity can lead to normalization and unquestioning acceptance of research practices. Do our practices and procedures as researchers become taken for granted? Do we need to stand back periodically and reassess our established routines and what we accept as ethical approaches? Implementing an unfamiliar recruitment process forced one of us to think again about recruitment. Issues of participant informed consent and researcher and/or funder power are ongoing issues and remain largely unresolved. As researchers we need to be sensitive to these issues, but as this study has demonstrated, participants are by no means passive. In addition, these questions also raise concerns about the lasting research footprint that our research activities might leave with participants, ourselves, and also fellow researchers. In this paper we now address some of these issues and begin to consider their implications.

Sensitivity and responding to emotion

Although ever present in qualitative interviewing, the mental health and employment project brought to the fore questions and ambiguities surrounding “appropriate” responses to emotion, whether and how to offer practical or emotional support, and the potential positioning of the researcher as therapist or “friend.” These issues span some key methodological debates and continue to be much discussed (Birch & Miller, 2000; Dickson-Swift, James, Kippen, & Liamputtong, 2006; Goodrum & Keys, 2007). The ambiguities and complexities were salient to both of us, and at times we felt uncertainty and conflicting emotions, and took different courses of action.

Although we recognized that all interviews are potentially emotional experiences for participants, we felt this to be particularly pronounced in this sensitive study area as in the interviews we asked people to relive difficult experiences in both their personal and work lives. Moreover, it was possible that the interviews could lead participants to confront or assess their own mental health condition in ways that they might not have wanted or needed to do in the past, thus reflecting on its impact on their past, present, and future lives.

One of us had some limited experience of interviewing people with mental health conditions as previous research with (in particular incapacity) benefits claimants meant that mental health conditions were sometimes an incidental part of study participants' circumstances. However, this was the first time her work had focused exclusively on people identified as having a mental health condition. For the other of us, mental health was a new area of work. We were concerned to conduct interviews appropriately and to be sensitive to the needs of research participants and wondered how best to equip ourselves in terms of interviewing techniques and responding to the experiences and views of those who took part.

As with any new topic of inquiry, we found it useful to consult the literature on other researchers' experiences alongside information available from specialist organizations. A practitioner-researcher working at the same institution was also approached as he had extensive experience and expertise in working with people with mental health conditions. An informal meeting was arranged, and the practitioner-researcher provided advice on what to expect when meeting people experiencing various types of mental health condition (for example, behaviors and medication effects) and ways in which fieldwork could be arranged to be most productive (for example, visiting people with depression in afternoons rather than mornings). As well as practical advice, the meeting provided a degree of emotional reassurance; we could share concerns with somebody who empathized as a fellow researcher but was also able to draw objectively on wider professional and clinical knowledge. Indeed, retrospectively, the author with little previous experience of interviewing people with mental health conditions recognized that unbeknownst to her, many participants previously interviewed might have had mental health conditions (current or past) that they had not spoken about in interviews. For her, this raised some uncomfortable personal questions about her own negative and rather naïve assumptions and/or fears about mental health, presumptions that the practitioner-researcher was able to put in context. This strategy of consulting with professionals and/or practitioners has also been used effectively by this researcher in previous projects; for example, speaking with social workers or teachers when embarking on research with children with disability. In all cases, it was felt to give the researcher greater confidence and enable a more sensitive approach to interviewing, which had benefits for both the researcher and research participants.

Moving on from issues of preparation for the interviews, there were also questions around managing and responding to emotion during interviews. Dickson-Swift et al. (2006) have noted that qualitative interviews share similarities with therapy as both are based on empathy and listening skills' giving space for participants to talk about personal issues to someone who wants to listen. However, there are also important differences. A therapist listens with the aim of helping participants, whereas a researcher may listen attentively but ultimately takes the information away, offering little in the way of feedback; in essence, the participant is helping the researcher. Moving toward a pseudotherapist role can be ethically inappropriate if researchers are not trained in this field and can leave interviewers feeling emotionally burdened and unable to provide appropriate support (Cotterill, 1992; Dickson-Swift et al., 2006).

Neither of us is a trained counselor, and we both recognized that there was no single answer to the question of how best to provide support when participants became emotional. Postinterview discussions between members of the research team revealed that what was perceived as appropriate at the time depended on the circumstances of each interview and the researcher's personal interpretation of this context while in the field. Goodrum and Keys (2007) and Kavanaugh and Ayres (1997) have advised researchers to be aware of and responsive to participants' coping strategies and management of distress, and to be led by participants at times of heightened emotion. This was reflected in our reactions as we employed a range of strategies to enable participants to express, address, or manage their emotions and, in doing so, to regain a sense of personal control. These strategies included allowing participants to set the pace and depth of an interview, allowing extended periods of silence (which for one of us was initially challenging as she perceived silence as a failure to develop rapport), moving on from evidently distressing topics, and facilitating ways for participants to take time out of the interview situation; for example, by changing the subject to a neutral theme; turning off the recording device, thus signaling to the participant that the research interview had been suspended; or physically leaving the participant alone for a time, for example by going to make a cup of tea. Reassuringly, this range of responses, which largely came naturally to both of us, bears close resemblance to research participants' desired responses to emotion during research interviews as reported by Graham et al. (2006).

Managing one's emotions as a researcher, which can involve allowing, acknowledging, and even integrating them to research (Holland, 2007; Hubbard, Backett-Milburn, & Kemmer, 2001), was also an issue that we faced. Both of us had personal and/or family experiences of mental health conditions and so carried with us varying degrees of personal “baggage” (Knowles, 2006) when conducting and making sense of our research interviews. For the researcher who had past personal experience of mental ill health, it had not initially occurred to her that the project might be challenging in this respect. However, early on in the fieldwork there was a period of uncertainty where it began to feel that the intense focus on the mental distress of others was unearthing emotions and reflections on personal experiences to an uncomfortable and unhealthy extent. This feeling passed, however, and she was able to reframe her past experiences as a useful context and contribution to her interest in the work rather than a problematic intrusion.

However, these experiences highlight the potential value of an outlet for researchers' personal emotions and reflections aroused through specific interview encounters. Many of the stories that participants shared with us were of traumatic and distressing events under sometimes shocking circumstances. At various times we experienced feelings of being overcome by the harrowing experiences of others and the need to unload these while being mindful of the confidentiality and anonymity assured to participants, feelings and considerations similarly recognized by other researchers (Bloor, Sampson, & Fincham, 2007). At our workplace there were informal opportunities to discuss fieldwork experiences with research team colleagues; however, no specific emotional support system had been built in to the study. This was partly because we had neither acknowledged nor anticipated the extent to which the project would affect us emotionally. However, it is very difficult to preempt the emotional impact of research encounters, a factor that Corden, Sainsbury, Sloper, and Ward (2005) have similarly noted.

Rapport and reciprocity

Past studies (Birch & Miller, 2000; Dickson-Swift et al., 2006) have demonstrated that researchers are often confused as to the role that they should take and the role that participants expect or want them to take. We experienced such conflicts and ambiguities in terms of the extent to which reciprocity of information was desirable and whether this was an appropriate means of facilitating rapport with interview participants. Again, each of us made different decisions on how to address these concerns at different times. Although recognizing the importance of rapport, we were mindful of what has been termed “doing rapport,” a situation whereby rapport is increasingly commodified and commercialized, with rapport presented as a skill that can be turned on and off as and when required (Duncombe & Jessop, 2002). Concerns have also been raised about the “ethics of empathy” (Holland, 2007); for example, that this conceptualization of rapport might lead to researchers using learned rapport skills in a manipulative manner to encourage participant disclosure, a situation that might lead to people sharing more than they might want to which postinterview they might regret (Birch & Miller, 2000; Birch, Miller, et al., 2002; Cotterill, 1992; Goodrum & Keys, 2007; Kavanaugh & Ayres, 1997). As researchers we thus need to balance effective investigative inquiry with sensitivity and respect for participant privacy and subjectivity.

Feminist literature frequently stresses the importance of the researcher developing an active relationship with participants based on trust and reciprocity (Oakley, 1981; Stanley & Wise, 1993). In contrast, others highlight the importance of professional detachment (Lupton, 1994) or role distancing (Goffman, 1961) to avoid role confusion and the potentially conflict-laden territory of researcher as “friend.” Adding to this complexity, research participants will have their own perceptions and expectations of the nature of the research encounter, and the role they confer onto the researcher. We both found that there were conflicts and uncertainties about how much to share with participants from our personal or family experiences in the interests of rapport building and offering reciprocity. During a small number of interviews we decided to disclose personal information in terms of either our own or close acquaintances' experiences of mental health issues. The effects of this disclosure both on the participants and on us were experienced differently in each instance, as the following examples illustrate.

On one occasion, to end the interview on a positive and more empathetic note, one researcher shared the success of someone she had known with a similar mental health condition after the interview had been completed. Assessing the impact of this information, she felt that it did achieve this end, with the participant interested in and apparently heartened by the progress of her friend. The other researcher had, at several points in the fieldwork period, reflected on whether it would be helpful or not to share her own past experience of mental ill health with participants. She had decided against this, personally unable to resolve the question of whether this would be genuinely helpful to participants or would appear a somewhat self-reflective opportunity for catharsis, which had no place in the interview encounter. However, in the final interview conducted, she did briefly share personal experiences, empathizing with a participant's description of the negative effects of particular medication. This disclosure provoked only the barest acknowledgement from the participant, had no apparent effect on rapport, and, indeed, left the researcher with the impression that in this context personal disclosure had been neither helpful nor appropriate, a feeling similarly echoed by Ribbens (1989):

These different experiences and participant reactions demonstrate that defining and maintaining research relationships is complex, especially when researchers are continually reassessing and making decisions about the encounter in progress (Dickson-Swift et al., 2006). Graham et al. (2006) have noted that in their sample of qualitative research participants there was “little evidence … of people wanting the interviewer to share personal information with them” (p. 31) and that advancing the researcher's perspective on the research themes could in some cases constrain the contributions of participants. Although both researchers were seeking to give back something to the person they were interviewing, this reciprocity of sharing what were essentially small snippets of personal information clearly is not, as Rubin and Rubin (1995) have noted, an “equal” exchange of information.

The research “footprint”

A final issue to arise from the mental health and employment project was that of providing appropriate follow-up support to participants; in other words, acknowledging and responding to what Graham et al. (2006) have termed the “footprints of research” left behind by the researchers' activity.

Central themes in the research interviews were people's experiences of struggling in work because of mental ill health and making the transition from paid employment to claiming benefits. People were asked to reflect on the extent and quality of support they received both for their mental health and for financial stability. Anticipating that some participants might not have knowledge of or access to effective support, we prepared information postcards showing contact details for local benefits and welfare advice services and for local and national mental health support services. These were taken to all interviews and given to participants who specifically mentioned a need for information or were offered when we sensed that this might be helpful. In practice, very few cards were distributed as participants were either aware of local information sources or did not demonstrate or express a need for information or support.

However, beyond this provision of practical and publicly available information, we both experienced dilemmas about our ethical responsibility to personally provide some form of postinterview support. Reviewing their research encounters and past studies in sensitive areas, Kavanaugh and Ayres (1998) have suggested organizing referrals for participants to support services and/or participant debriefing and that researchers might employ follow-up procedures such as telephone calls. Some feminist researchers have also asserted that there is a moral obligation for researchers to provide postinterview support (e.g., Brannen, 1993). Team discussions for the mental health and employment project concluded that follow-up calls as a matter of course would not be prudent in this instance, in part due to the limited therapeutic or practical support that we were able to offer should this be required.

However, one of us felt that merely handing out printed information cards was inappropriate to the emotional distress that two participants had shown during their interviews and would not be viewed as helpful by the participants. A more personal approach was considered necessary, so the researcher telephoned the two participants later on the day of the interview to check how they were feeling. At the time, this seemed the best and most compassionate thing to do. However, retrospectively, the appropriateness of this response was a source of personal conflict, as the researcher recognized that potentially problematic or detrimental situations could have emerged for both her and the participants.

Although steps had been taken to minimize any negative or distressing footprints for the research participants, there remained unresolved questions for this researcher about whether she had done the right thing. Would the participants who were telephoned have viewed this call as intrusive or unnecessary, or an infringement on their personal space? Would it have been viewed as tokenistic, given that the researcher could not actually change their situation or provide any real support, practical, emotional, or financial? Was making contact after the event an inappropriate crossing of the personal-professional boundary? For example, as Cotterill (1992) has suggested, volunteering help could be viewed as presumptuous or patronizing by participants. Furthermore, there was the question of whether the researcher's motivations for making the call were truly altruistic or whether she was making the call primarily to ease her conscience. This could have been a dangerous situation to enter as she did not have a clear plan of action if there were problems. Questions also remained for both of us as to whether a selective, individualized approach to follow-up support was most ethically appropriate. Should all participants have been offered information cards whether there was an apparent need or not? Were there participants whose need for support went unnoticed? Moreover, were there people who had not displayed distress during the interview but who later began to feel differently about what had seemed a comfortable encounter at the time, experiencing delayed reactions of distress or regret at taking part in the study?

Beyond concerns about the footprint we created, both of us were also struck by the impression left on us by meeting with and hearing the experiences of participants. Despite having conducted numerous interviews with people in challenging or disadvantaged circumstances, one of us felt a previously inexperienced depth of emotion on coming away from these encounters. Writing thank you letters to participants, often personalized with a relevant line or two referring back to the participant's individual circumstances, is a common practice in the research unit. However, in the mental health and employment study this researcher often found herself moved to write extended passages of thanks and good wishes. Again, personal-professional boundaries became somewhat blurred, and the researcher's subjective and individualized correspondences perhaps raise questions about the desirability and appropriateness of such personal responses in “professional” research activity.

Although the research was not conducted specifically within a feminist paradigm, it is perhaps relevant that both of us are female and that the interviews that had stirred the strongest emotional responses or desires to follow through with some kind of support offer were predominantly those with female participants. As Ribbens (1989) has noted, there is a close and reciprocal relationship between empathetic listening, care, and intimacy, and as women we perhaps “expect to nurture those that [we] have listened to so carefully” (p. 587) in qualitative research interviews. Being compelled to detach ourselves from research participants at the end of the interview encounter for methodological, ethical, or practical reasons at times did not sit comfortably with our instincts as women or simply as compassionate social beings.

Finally, as the research project moved on, there were also significant differences in our lasting experiences of the fieldwork. One of us was assigned to the project for its duration and was able to immerse herself in the data set, becoming more familiar with participants' stories, working through them in analysis, and endeavoring to arrive at conclusions and implications that might in some way improve circumstances if not for the participants specifically then for others in the future. However, the other was assigned only to the fieldwork stage and was thus not afforded the opportunity to stay connected or develop a deeper ongoing relationship with participants in this way. Discussion of postfieldwork experiences with another team member raised the issue of “unfinished business”. For the researcher with limited involvement, this discussion elucidated her feeling that the emotional engagement and intensity of the fieldwork period had in some ways been left unresolved. At the start of the project she had presumed, in contrast to other studies she had worked on, that a more detached role would be adopted because of minimal involvement in research design and planning: This was not her project. It was, therefore, a shock when she felt emotionally moved and almost responsible for some interview participants, and this was harder to deal with because of limited postfieldwork project engagement with the absence of opportunities to work through personal emotions via processes of analysis and report writing and pressure to quickly move on to other projects and priorities. A lack of connection to the final study output also left a sensation of not being able to give anything back to study participants.

Discussion and concluding comments

In this paper we have explored a range of issues that emerged as we planned, conducted, and jointly reflected on our experiences of in-depth interviews with people who had claimed an incapacity benefit as a result of mental ill health. Despite our both having substantial previous experience on a number of varied qualitative research studies, this particular project gave renewed focus to a number issues and concerns, including consent, emotions, rapport, and support. We have discussed how we approached these issues and have attempted to demonstrate how each issue has relevance and potential implications for the well-being and ethical requirements of both study participants and researchers. In this final section we summarize the key lessons that we drew from this experience and pose some ongoing questions that we hope other researchers identify with or will find useful to consider.