Ethical Dilemmas of a Clinician/Researcher Interviewing Women who have Grown up in a Family Where There was Domestic Violence

Abstract

This paper attempts to illuminate some of the ethical dilemmas of a clinician/researcher interviewing women about a sensitive topic- their experience of having witnessed domestic violence in the family they grew up in, as part of a grounded theory study. Vignettes are presented to illustrate the self-reflexive process of the researcher and how she understood the effects of the interview process on her and the participants. The authors argue that doing in-depth qualitative research interviewing is an intervention in the life of the participants, especially, but not only, when the researchers are clinically-trained. However, this clinical training may also be an important resource from which to draw from, to act ethically and understand some of the complexity of the interaction between researcher and participants.

Introduction

The ideas in this paper developed as a result of the discussions that we, the authors, had about the ethical dilemmas that arise when interviewing participants on sensitive topics, during the research supervision process of a grounded theory study (Charmaz, 2006; Glaser & Strauss, 1967). The participants were eight adult women who grew up in a family in which they witnessed domestic violence. They were interviewed around their understanding of how these experiences have impacted their adult relationships, particularly their significant relationships with partners and their children, if any. This paper also focuses on research “ethics in practice” (Guillemin & Gillam, 2004, p. 262) – on how ethical decisions have to be made at different steps in the research process (Hallowell, Lawton, & Gregory, 2005), illustrating how ethics in research goes beyond gaining approval of the institutional research ethics committee boards.

The ethical dilemmas discussed in this paper also center around the issue of the researcher also being a trained clinician (a psychologist and systemic psychotherapist) and on some of the dilemmas that arose, precisely with managing the dual roles of being a researcher and a clinician. Similar dilemmas are also experienced by researchers trained as nurses who interview people with HIV or with terminal illness (Eide & Kahn, 2008; Kylma, Vehvilainen-Julkunen, & Lahdevirta, 1999). We would like to show how doing in- depth interviews around sensitive topics leaves an impact on these participants and on the researcher. Having access to clinical supervision or clinical training may be an important resource for researchers in this context, in that it may help them make sense of their interactions with the participants and reflect on the possible therapeutic or distressing effect the interview may have on the interviewees and on the researcher.

In our supervision discussions, we also wondered about how having a clinical background impacts the actual interviewing process and about the difference if any, between a clinician researcher and an academic researcher. Would this make a difference to the participants? In this context, is it more likely that the boundaries between therapy and research become more blurred - hence raising issues of what participants have consented to? In addition, how does having a clinical background impact the reflexive process of the researcher to make sense of her/his interactions with the interviewee? Does this help the researcher become more sensitive to issues of harming or causing distress to one's participants?

These and other dilemmas will be addressed in this paper and illustrated through the use of vignettes. The use of the personal pronoun “I” will make reference to the researcher's reflections while the use of “we” in the paper acknowledges the iterative process between the supervisors and the researcher. Before we discuss some of the ethical dilemmas we discussed, we will outline the philosophical assumptions and other theories that form the backdrop for this study.

Philosophical Assumptions Underpinning the Study

This research study is informed by the tenets of Cooperative/Collaborative Inquiry, first proposed by Heron (1971, 1996), expanded by Reason (2003) and further developed by Gehart, Tarragona, and Bava (2007). According to this framework, influenced by a social constructionist stance (Gergen, 1999), research is done “with” rather than “on” the participants. Moreover, the focus is on the process of the researcher and participant creating together knowledge and meaning, in dialogue. This highlights the assumption of inter-subjectivity between researcher and participants and the mutual creation of data, which are concepts common to qualitative scholars skeptical of researchers claiming to be uninvolved observers in the research process (Schwandt, 2000). Another characteristic of collaborative inquiry is that researcher and participants are seen as inquiring together. Thus, participants are seen as having the capacity to reflect and be co-researchers and as such are involved “in defining what questions need to be asked and identifying processes that might be useful in answering those questions” (Gehart, Tarragona, & Bava, 2007, p. 375).

Our study also acknowledges resonance with feminist interviewing processes (Hesse-Biber & Leavy, 2006) that share an ethical concern about the power and authority of the researcher and the uncomfortable position of getting data from participants as a means to an end (Olesen, 1994). This highlights some of the tensions of doing research within a collaborative framework. In fact, feminist researchers such as Oakley (1981), Wilkinson, (1988), Stacey (1991), and Reinharz (1992), have also written about the difficulties that challenge the researcher to empower participants in the research process.

One way of addressing the hierarchy between the researcher and the researched is for the researcher to know and make explicit his or her own position as a researcher – one's social background, training, one's personal and professional beliefs, assumptions and feelings and how these are affecting the different steps of the research study. This process can be termed as researcher's reflexivity (Finlay & Gough, 2003).

Reflexivity in Qualitative Research

Historically reflexivity has long been explored in qualitative research and there is a wide array of definitions depending on one's theoretical position and one's discipline (Roulston, 2010). There are also varied views about its usefulness in social research. For us as researchers, critically examining the interviewer's interactions in research interviews is one way of acknowledging the researcher's subjectivity and it is also a way of learning more about ourselves as co-constructors of data. Moreover, in Shaw's (2010) words, a reflexive process in research helps “proactively manage myself in my interactions with my participants, and the world and to actively explore how these encounters impact my pre-existing beliefs and knowledge- my fore understandings- in order to understand afresh the phenomena I am studying” (p. 241). This definition echoes what we understand by reflexivity. It also stresses the significance of the interpretative element in this process, outlining what Finlay (2003) refers to as hermeneutic reflection. Engaging in a reflexive process can also be a way of increasing the integrity and trustworthiness of qualitative research, which is also a component of doing ethical research. However, this aspect will not be dealt with, in this paper.

Most qualitative researchers will agree that doing reflexive research and actually writing about it is not easy. One can be accused of doing too much “academic navel-gazing” (Finlay, 2002, p. 215), of privileging the researcher's voice over that of the participants, which can also be understood as a form of Othering when conducting research with people (Wilkinson & Kitzinger, 1996). Moreover, our reflexive accounts are themselves also influenced by our social contexts and hence are but one way of seeing the world, from a particular perspective. This means that our understanding is in itself our construction and may be criticized as being a too simplistic rendering of the research encounter.

Writing reflexively may also induce some anxiety at exposing one's self and one's ethically challenging interactions with research participants. At the same time, we believe that these anxieties and challenges need to be shouldered with the responsibility that comes with doing ethical research, especially on sensitive topics. Recent publications such as Sinding and Aronson (2003), Shaw (2010) and Garfield, Reavey, and Kotecha (2010) have in fact ventured down this path and have produced enlightening texts that illustrate how reflexivity can become embedded in qualitative research.

Literature about Interviewing Around Sensitive Topics

The increase in research on women's experience of interpersonal violence in the 1990's (Kruttschnitt, McLaughlin, & Petrie, 2004) has generated with it concerns about harming trauma survivors by causing them emotional distress. Studies that have asked the research participants about their distress to different trauma assessment procedures (Carlson et al., 2003; Griffin, Resick, Waldrop, & Mechanic, 2003; Hlavka, Kruttschnitt, & Carbone-Lopez, 2007) suggest that whilst some respondents in these studies did report unexpected emotional upset from victimization research, these studies over all indicated that research on victimization seems well tolerated by participants and “adverse reactions appear less common than previously anticipated” (Hlavka et al., 2007, p. 898). Another more recent study whose authors conducted a qualitative study on how participating in in-depth interviews impacted survivors found that the majority of the participants experienced the interview to be “helpful, supportive, and insightful” (Campbell, Adams, Wasco, Ahrens & Sefl, 2010, p. 60).

Furthermore, participation in research studies has actually been found to be therapeutic (Gale, 1992) and of benefit to those who participated. Participants have experienced a decrease in intrusive thoughts over time when compared to a control group (Lutgendorf & Antoni, 1999) or felt that they learnt more about themselves (Reich & Kaplan, 1994). Similar effects were reported by participants in a study by Owen, Heyman, and Smith Slep (2006) who looked into the impact of participants completing questionnaires (Couples Conflict Protocols), engaging in conflict conversations and being interviewed individually about anger escalation and de-escalation during these conversations. Both men and women rated the impact of the study as helpful to them personally and to their relationships.

Besides looking at the possible distressing experiences and helpful experiences for the participants, one also needs to reflect on one's position with respect to these participants – as to whether seeing them solely as “victims” in need of protection, involves perpetuating their victimhood. Campbell and Adams (2009) asked participants why they have chosen to participate in a qualitative study about rape. The reasons given were to “help other survivors, (b) to help themselves, (c) to support research on rape/sexual assault, and (d) to receive financial compensation” (p.395). Such findings also help one to think about the resilience of trauma survivors. Not taking into account that altruism and a sense of agency may be other important factors for participants taking part in sensitive research, could be, in effect, disempowering to these women.

Trauma Theory

In our conversations, we discussed how the above literature albeit insightful, does not explain how and why distressing and positive experiences may occur in the research interviews. We will now draw from the trauma literature to suggest possible explanations.

Numerous authors such as van der Kolk and van der Hart (1991) and Terr (1991), postulate that traumatic memories are different than memories of ordinary experiences. When we speak of memory in our everyday language, we refer to the story that we have of our life and what we tell others. This is usually referred to as “explicit memory….[and] includes semantic, episodic, narrative and autobiographical memories” (Jonsson, 2009, p. 440). It is what is at work when our recall involves words, such as a description of an event or a list of instructions. This memory system is usually not completely functioning until a child reaches her or his third year.

Another important memory store called the “implicit memory system” (Rothschild, 2011, p. 42) comprises the emotional, somatic, physiological and behavioural part of our experiences. This is a non-verbal system, automatic and mostly unconscious. Usually the two memory components are integrated so that when we remember something, we remember the story and the affect associated with it.

When a person is faced with an affectively overwhelming experience, as happens in very stressful and traumatic events, the explicit memory is usually inhibited because it is sensitive to stress hormones. This may result in the storage of the memory as a feeling or a sensation only, without the necessary explicit facts or progression of an event (van der Kolk & van der Hart, 1991). Thus, a person may have very strong feelings of terror or pain or vivid recollections of an event and lack an explanation as to how she or he came to these feelings or memories, which in turn leaves this person as baffled and confused as the person was in the original experience. In addition, the intense arousal and possible dissociative responses (a disruption in a person's normal consciousness or psychological functioning) may also lead to a fragmentation of the experience and a loss of recollection of the traumatic experience (van der Kolk, 1996). These memory gaps, in fact, may hinder survivors' making sense of their experiences, and persons' fragmented narratives have been taken as possible indicators of them not having yet developed a clear understanding of the trauma (Foa, Riggs, Massie, & Yarczower, 1995). At the same time, survivors of trauma report that as they became aware of different elements of their experiences, they slowly start to construct a narrative that helps explain what happened to them (van der Kolk, 1996).

This same author explains that under ordinary conditions, survivors often have a fairly good psychosocial adjustment and cannot induce the return of traumatic memories at will. However, high states of arousal or anxiety seem to promote retrieval of traumatic memories, or sensory information or behaviours associated with the traumatic experiences.

It is presumed that when participants show an interest to share their experiences as survivors or witnesses of abuse as part of a research project, they are already in possession of narrative (possibly in construction) and have a related understanding of what happened to them. The emphasis on making meaning of their experiences as a result of their research participation can also aid them further in the processing of their experiences. In fact studies by Pennebaker and Francis (1996) have shown that the use of causal and insight words such as “because, realise” indicate efforts to understand traumatic events and a greater use of these causal words in narratives have been related to lower post treatment anxiety. This finding also supports the studies that show that greater understanding of traumatic events is related to better outcome in physical health (Alvarez-Conrad, Zoellner, & Foa, 2001).

At the same time, the process of understanding is facilitated in the context of a working alliance between interviewer and interviewee that is supportive, empathic, and compassionate. This is also emphasized in the literature about the therapeutic alliance between therapist and client, which has been identified as the most important determinant in treatment continuity and success (Beutler & Harwood, 2002; Horvath & Symonds, 1991) and has also been shown to be predictive of positive outcome (Bourgeois, Sabourin, & Wright 1990; Gaston, 1990). In the case of a research interview, the following quote by a participant in an interview research on rape survivors illustrates the above:

You were compassionate. You were understanding and you knew when to laugh. You knew when to be serious and you don't rush people through it, like hurry up, come on. You took your time. I think that you did a good job. If I'd just been raped last year, I'd feel very comfortable coming to you. (Campbell et al. 2010, p. 76).

For many of these participants, the experience of having someone listen with patience and real engagement was in fact therapeutic. They reported that it was very helpful to have someone listen to them and really hear what they were saying, without interrupting to give advice or pass judgements. What also made the experience positive for these survivors were the new insights and connections that resulted from the research encounter.

Whilst providing such a supportive context, the sensitive research interviewer also needs to give paramount importance to the participants' well-being and stop the interview if at any point the interviewee becomes particularly distressed. The interviewer would need to empathize with the participant and discuss with her whether she wanted to continue. In line with the trauma theory discussed above, this would help to avoid situations of high arousal in the interview which would possibly induce a re-traumatizing experience for the participant. It is also important to spend time talking about the experience of the interview and how it impacted the interviewee (debriefing) at the end of the interview so as to help the interviewee back to a normal state of arousal from the intensity of the interview. We would argue that debriefing needs to take place after all sensitive research interviews and not only if the interviewee becomes upset.

The below vignettes will highlight the application of some of the above theory and will attempt to illustrate other ethical challenges that the researcher experienced in her research. The following section will outline the methodology and method of the study

Research Design and Method

The effects of domestic violence on children have been well-documented in the extensive literature that has been published in this last decade. Several research studies on child witnesses to domestic violence have concluded that children's exposure to parental violence is associated with a wide range of psychological, emotional, behavioural, social, and academic problems. (Fantuzzo & Lindquist, 1989; Jaffe, Wilson, & Wolfe, 1986; Margolin & Gordis, 2000; Wolak & Finkelhor, 1998)

These studies indicate that exposure to family of origin domestic violence has long- lasting effects on the psychological adjustment of child witnesses and exposure also impacts their social adjustment and emotional development. Researchers indicate that the relationship between exposure and possible impact is complex and that “there is no easy way to assess the direct impact of one traumatic event during childhood on adult symptomatology when other closely related family characteristics may be present that may produce a further impact on adult functioning” (Henning, Leitenberg, Coffrey, Turner, & Bennett, 1996, p. 47) . This statement thus highlights the difficulty that quantitative researchers have when trying to isolate the effects of witnessing physical parental conflict from the effects of other related factors.

A grounded theory methodology was chosen to focus on the processes, the interacting and mediating factors that are involved when witnesses of family-of-origin domestic violence negotiate their relationships, particularly their intimate relationships with a partner and their children. As Charmaz (1995) and Pidgeon and Henwood (1996) argue, this approach is particularly appropriate for research topics that are contextually-sensitive. This approach was also chosen as it was designed to help ask participants how they understand the experience of the process (in this case living in a home where there was violence and how it has affected their relationships as adults). This is the rationale for the methodology chosen.

Recruitment

One of the ethical principles that was foremost in our mind whilst planning to interview the participants around their memories of witnessing domestic violence in the family that they grew up in was that of responsibility (British Psychological Society Code of Ethics, 2009), specifically referring to the Standard of Protection of Research Participants. We were aware of the need to put into place various strategies that would minimize distress for our informants. For this reason, we made the decision that the researcher would interview adult women who were in paid employment and who had access to supportive networks. This was informed by the growing body of literature in the social psychology field that stresses the importance of social support in the reduction of stress and the improvement of well-being (see Sarason, Sarason, & Gurung, 1997; Umberson, Chen, House, Hopkins, & Slaten, 1996).

These women were recruited through professional colleagues, mainly psychologists and family therapists who had worked with the prospective participants in the past or were presently working with them as their clients. The professionals were sent the interview schedule so that they could assess whether, in their opinion, the women they had in mind would be suitable participants for the purpose of the research but also to evaluate the present emotional state of these women, if they were working with them in the context of therapy. As the referring professionals, they also agreed to support the referred participants if the women felt that they needed a follow-up meeting after the research interview. This was not agreed to through a formal written contract but through a verbal agreement with the researcher in the spirit of collaboration and collegiality.

Once the potential participants indicated their willingness and interest to the researcher's colleagues to be approached, the women were asked if they had discussed the interview schedule with the referrer and if not, they were sent the schedule before the interview, so that they could see it beforehand and reflect on it. This was done to give them more control over what they would talk about. They also had the opportunity to address any queries with the researcher before the actual interview through email or a phone call depending on their preferred means of communication. One woman at this stage declined to consent to be interviewed, realizing that participating in the interview was going to be stressful for her.

The Data Collection Process

The Pre-Piloting Stage

During the interview with the first three participants, the interview guide (see Appendix A) was pre-piloted in terms of the structure of the interview and also the content of the questions. The participants were asked whether they would feel more comfortable with an unstructured interview where they would be given space to tell their story in a way that fit for them or whether they preferred our conversation to be guided by the questions in the interview schedule. This was done for methodological purposes given that the research topic is under-theorized but also to create a context of “safe uncertainty” (Mason, 1993). The concept of “safe uncertainty” acknowledges both the degree of risk that the participants might feel that they have to face in speaking about their childhood experiences as well as the participants' sense of control on the interviewing process that helps them to feel less overwhelmed. The participants were also asked which questions they would find comfortable to answer and which questions they would feel less comfortable with. In addition, they were asked about the sequence of the questions, and which questions they thought would be useful to answer first.

All the participants in the pre-pilot stage of the research preferred to be guided by a semi-structured interview guide, feeling that they would be quite overwhelmed without this “structure”. This led us to hypothesize that the interview structure probably helped in the organization of their experience, keeping their anxiety low and helping them feel less fragmented and more able to think (referring back to the traumatic impact on memory systems discussed earlier). The participants also felt comfortable with the questions and with how they were articulated. In addition, they felt that it was important to tell their stories from the beginning, that is, talking about what they went through in their childhood as this is how they thought that I could understand their present relationships. All the women at pre-piloting stage and during the actual data collection process were interviewed once, and each interview (see Appendix B for interview schedule) took an average of one hour to one hour and a half.

Ethical Dilemmas in Practice

The feedback from the participants at the pre-pilot stage was encouraging, in terms of the minimization of risk to the interviewees. At the same time, these participants were primarily going to be interviewed as part of the researcher's doctoral research and therefore for the purpose of career advancement, which raised the issue of the potential for exploitation of their experiences for the researcher's benefit. This theme was also kept in mind during the actual interviewing process, so as not to encourage participants' self- disclosure, possibly about areas that they did not anticipate talking about.

Systemic practitioners and other clinicians may have an advantage as researchers in that they are trained interviewers and have experience in listening, following feedback, focusing on meaning, and helping the interviewee to manage affect–as has been mentioned earlier–which are all helpful skills when exploring sensitive areas of a participants' life. However, it is also important to be aware that these skills are not misused to encourage participants' self-disclosure, possibly about areas that they have not consented to talk about (Kvale & Brinkmann, 2009). These concerns perhaps become more pertinent when qualitative researchers use semi-structured interview guides as an interview format which implies, at times, discussing topics which have not been consented to.

During the interviewing process, it was helpful to be continually mindful of these different issues. Thus, during the interview I would summarize what the participant had disclosed in order to acknowledge my attending to their story and then ask if I could take them back to the questions in the schedule. In the debriefing stage of the interview, I asked the participants how they felt about the fact that I was a psychologist and a systemic practitioner and also a researcher. They reported that it made it easier for them to talk about their experiences precisely because they knew that I would understand their experiences; that they were not “freaks”; that I would not judge them and that I also would not become overwhelmed by their self-disclosures because I was familiar with similar stories through my clinical practice. It would be interesting to pursue this feedback further through future research, where participants could be asked about their views of being interviewed by clinician/researchers and asked to describe any helpful and unhelpful experiences.

Dowd & Wilson (1995) stress that informed consent should not be viewed as a one-time process in the beginning of the study but that it should be a continuous communication process. This is important particularly in grounded theory studies owing to the emergent design nature of the study where neither participants nor researchers are fully informed at the beginning how the study will evolve and what kind of data will be gathered (Kylma et al., 1999). Therefore it cannot be assumed that consent at one stage of the research is still valid at another stage and participants need to be given information about the different stages of the study and their consent asked more than once. In this study, since the participants were interviewed once, signed consent (Appendix C) was asked and given once but the women were given various opportunities to ask for more information and to consent or not, to participate in this research at referral stage, at contact stage, and at the interviewing stage.

Emotional Impact of the Interviews on the Researcher

In qualitative research, there is now increasing recognition that undertaking sensitive research not only necessitates an assessment of the risk to participants but also requires a consideration of the risk to researchers (Lee-Treeweek & Linkogle, 2000; McCosker, Barnard, & Gerber, 2001). The “Code of practice for the safety of social researchers” (Social Research Association, UK, 2001), which outlines in detail potential risks to personal safety and procedures to reduce these risks, may be an important resource for the qualitative researcher. Other writers have discussed the emotional risks of doing sensitive research, for example, experiencing feelings of guilt and vulnerability (Dickson-Swift, James, Kippen, & Laimputtong, 2007), having episodes of insomnia and nightmares (Cowles, 1988), and experiencing emotional exhaustion leading to researcher burnout (Dickson-Swift et al., 2006).

In the following vignette, I would like to discuss the kind of emotional impact that I experienced as a researcher and how receiving supervision was essential to help me process the interviews and help me deal with the intensity of emotions. Supervision was given by the lead supervisor and the co-supervisor, both clinical psychologists and family therapists with extensive clinical experience in the field, including one of them having set up and managed a domestic violence service.

Despite the fact that the difference between therapy and research was clearly made explicit at initial contact and also prior to the interview, and in spite of the fact that I do have clinical experience working with traumatized clients, I was taken aback at how emotionally draining and demanding some of the interviews were, especially the initial interviews both for me and for my participants. The distinction between therapy and research reflect the different goals of the therapeutic and researcher encounter–one depends on longer term contact and is meant to facilitate change and the other mainly takes place to gain information. This distinction was clear to me and I also attempted to explain it to the participants, discussing how we will only meet once and how their psychologists would continue working with them on any issues that might come up for them during the research interview. With hindsight, I realize that I expected to be less emotionally impacted by this one-time encounter. However, the experience proved otherwise.

Rebecca ¹ was one of the first women that I interviewed. She readily engaged in the interview process and each question yielded stories about her childhood experiences of witnessing her father's aggression against her mother and also being physically abused by him. Rebecca's interview had not been her first opportunity to talk about her experiences. She had been to therapy for a long period of time and she told me that therapy had helped her understand and process aspects of her traumatic experiences. Thus, I felt safer that there would be a lowered risk of our interview triggering unresolved emotions related to what she went through in her childhood. At the same time, I was watchful of her process within the interview itself.

Rebecca's father not only was violent to her mother but also very physically abusive towards her. She described several descriptions of the raw, sadistic violence of her father's beatings, for example, beating her with chains, or throwing a knife at her and the knife getting stuck at the point of her breast bone. In a cruel twist, he also used to demand that she judge who was right and who was to blame between him and her mother. Rebecca then used to get very angry at her mother for not being able to stand up to him and not helping her (the daughter) in the judgement process.

Not only were these narratives of callous cruelty and instrumental violence but Rebecca also described these experiences in a fragmented manner which made it difficult for me to follow her story. The lack of coherence left me wondering about the level of processing, despite the fact that she had been in therapy for a long time. I was torn between bearing witness to her suffering as a child at the mercy of an abusive father and deeply empathizing with her experience and also feeling shock and horror at what a father inflicted on his daughter. I was also aware of my feelings of anger towards him as the perpetrator. There were instances where I “spaced out” and was unable to hear her. In addition, the incoherence and fragmented aspects of the story were at times difficult for me to hold together and I felt I needed to stop her at times to ask her questions to be able to try and piece together what she was saying.

After the first hour and a half of our interview, I was aware of feeling fatigued and I was also conscious of the fact that we had covered only about half of the interview material. I was also aware that Rebecca had agreed to being interviewed for about an hour or so and I did not want to make excessive emotional demands on her. I made the decision at that point to ignore the rest of the questions and to try and find the right opportunity to ask her about what she felt about starting to wind down our conversation, privileging her emotional well-being and also being conscious of my emotional exhaustion.

I would also like to present another vignette which highlights a similar but slightly different emotional impact on the researcher. In her interview, Jane described her happy childhood memories with her siblings and mother. Jane's mother was murdered by her father. To my relief, I felt more able to be with her despite the sadness of her story than I had been with Rebecca. Jane's narrative was also more coherent; however, after the interview, it was quite difficult for me to shake off the immense sadness that I felt as a result of hearing her story. With Rebecca, the horror and grief were with me in the research interview; but, I carried Jane's interview with me after. This participant's interview took place in the evening and after the interview, I wrote down my thoughts and feelings about the interviews (Pennebaker & Graybeal, 2001) in my research diary to help me process some of the traumatic material, and in an effort to purge myself of some of the emotions. Nevertheless, I ended up dreaming about Jane and her family, an experience which I found quite distressing.

Such experiences highlight the need for qualitative researchers to leave enough space between interviews to process any information that may be distressing, not to have any interviews when one is tired after a long day's work and to have prompt personal supervision where researchers can get the opportunity to debrief and to disentangle some of the emotions that were experienced.

Before discussing the supervision process, I would like to conclude this section with what seems like a shift in the emotional impact on the researcher. It seemed that as I conducted more interviews, I was able to be more present and actively in tune with the emotions of my participants and the themes of their experiences. It seemed that there was a development in my ability to stay with the pain of the participants' experiences, almost like an extension of my empathic ability (Wilson & Lindy, 1994). This process seems somewhat similar to what Bar-on (1996) describes in his own reflections of interviewing children of Nazi perpetrators in Germany.

The Supervision Process

The relationship I have with my supervisors is characterized by a trusting, respectful relationship, where there is emphasis on collaborative practice (Goolishian & Anderson, 1992) and the cooperative, mutual generation of ideas (Anderson & Swim, 1994). This relationship has developed through the years working together in different professional contexts. In supervision that is influenced by social constructionism, it is a given that the supervisors usually have more knowledge, skills, and experience to draw on but this expertise is used to facilitate the supervisee's progress. Issues of hierarchy, expertise, and “truth” are as much as possible talked about rather than left as implicit or vague concepts. This helps to further foster an atmosphere of trust and respect which encourages supervisees to share their mistakes and dilemmas.

I have found it very helpful to share with my supervisors my research journal entries after each interview as well as for them to comment on my interview transcripts and memos. Hearing my supervisor's remark on the horror of the sadistic violence that the participants recounted or share the distress that they felt on reading a particular transcript helped normalize my experiences of distress. Drawing from literature about therapists working with trauma (Pearlman & Saakvitne, 1995), we talked about how powerful feelings are evoked in therapists when working with traumatized people and these same feelings seem to be evoked in interviewers too, who listen to the same traumatic material. One can feel overwhelmed; feel the dread, anger, shock, and horror at some of the atrocities that people have to suffer. These powerful feelings can also evoke some of our unacknowledged feelings of our own childhood experiences. Faced with this emotional roller coaster, one can react by withdrawing emotionally (Herman, 1997) or become overly sensitive to responding, identifying with the participant, and over-worrying about issues of safety and harm.

Having the opportunity to discuss these issues in the safety of the supervisory network helps create a thinking space–a “stable third” (Cooper & Vetere, 2005, p. 17) to explore these tensions. Another point of view in understanding the dynamics at play is offered which helps in managing the anxieties and ethical dilemmas that emerge whilst doing this kind of research.

Privileging Participants' Well-Being over Completing the Research Interview?

One of our supervision discussions involved the transcript of a particular participant–Roberta. During the line by line coding of my transcript, I realized that I had not pursued some of her feedback to clarify issues and to build the interview conversation on her feedback. A number of research hypotheses remained unanswered. It seemed that I, as a researcher was less present in the interview and my contribution in the conversation was rather absent when compared to the other interviews. In supervision, we discussed different possible explanations as to what was going on with this participant. A short vignette will hopefully throw some light in understanding some of the dynamics at play.

When I started the interview with Roberta, she found it very difficult to remember what life was like at home when she was young. She could not remember as far back and what she could remember were what seemed to me, very vague memories. She also could not remember much about her relationship with her mother and what this was like when she was young. Interpersonally, I also got a sense of how much the participant felt alone and lonely in her present life and she confirmed this later on in her interview.

I was also very much aware that she still lived with her parents and that after the interview she had to go back to her family and her abusive father. I wondered whether Roberta could not remember because she had repressed some memories as a result of going through some traumatic experiences and I found myself being very cautious not to be intrusive in my questions. Roberta's process could possibly be understood as one characterized by deactivation strategies (van der Kolk, 1996). This is when a person dismisses emotions and painful memories and presents a facade of calm and security but hides an internal distressed state, which still needs to be noted and the participant helped to come back to his or her own sense of composure and feeling of “being together”. With regard to the questions in the interview guide, I felt that especially in her case, I needed to give control over to her. I asked her to pick and choose which questions she felt able to give me feedback about. She preferred that I would guide her through the interview guide, which I did. But when she presented with what to me seemed like vague answers, I did not pursue with further prompts.

In a sense, one can argue that my cautiousness led me to prioritize the participant's needs perhaps comprising the research interview. The narrative about trauma and traumatic memories possibly constrained me from thinking about other factors which could have been going on in this interview. In a discussion with a researcher colleague, we wondered whether apart from the notion of traumatic memories, other factors were at play, for example, that Roberta was having difficulty with the actual wording of the questions. At that point in time with Roberta, in fact, I did not inquire about this. Perhaps, with hindsight, it would have been helpful feedback to the research process had I asked her about this.

This vignette attempts to highlight the importance of building in pauses for processing and reflection within the interview itself so that the researcher can think about different explanations of what could be going on in the room at that point in time, and then, have the presence of mind and the skill to use immediacy to get feedback from the participant about what might be hindering the interviewee from answering the question. This also assumes that the participant would also be given the option of not answering the questions if this is what she would prefer.

Responsibility to Avoid Harm and After Care

According to the British Psychological Society's code of ethics (2009), researchers at a minimum should ensure that at the end of the interviews, participants are debriefed to ensure that any emotional distress during the interview has decreased. As mentioned earlier, debriefing is an important step for all participants in sensitive research and not just the ones who have experienced distress during the interview. Participants also need to be aware of where they can access additional emotional support if needed. In this study, the participants were reminded that they could contact the referrer–the psychologist or psychotherapist who had talked to them about the study in the first place and if they were in therapy, I encouraged them to discuss some of the issues discussed in the research with their therapist to ensure continuity of care. In the context of the debriefing process, I also obtained the participants' verbal consent to contact their clinician as part of giving feedback on the emotional state of the interviewee and on how the interview went. This professional contact did not involve disclosing details about what went on in the research interview but it was part of a process of feeding back to the referrers any distress that the participants might have felt as a result of the interview, to ensure that continuity of care takes place.

The Research Interview as a Helpful Experience to the Participants

As the interviews progressed and I could personally experience that these particular participants had not been harmed but had actually told me that they felt that the interviews had been helpful to their growth, I found myself becoming a little more courageous and less afraid.

In the following two anecdotes, two of the participants discuss how the research interview was a helpful process for them. Certainly it would be interesting to explore further this theme in a separate research project as has been indicated earlier.

Sandra, a woman in her early thirties was asked by her psychologist whether she would be interested to participate in this study. She was an intelligent woman, who was successful in her career but had difficulties in her interpersonal life. During the interview, she presented herself as being very reflective and readily shared her thinking with me. After the interview, she told me that she had found the interview very useful, in that she had never before made any links between her past experiences and her present. Even when her psychologist had asked her whether she was interested to participate, she had been surprised that her psychologist had identified what she went through in her childhood as domestic violence, so she came to the interview curious as to what would emerge.

About a week after the interview took place, Sandra sent me some of her written reflections through email. She re-iterated that although her knowledge on domestic violence was vast (as a result of her job), she still found it hard to believe that what she sees at work, is something that she has lived through. But after the interview, she could see how the physical, mental, and verbal abuse that she had suffered left their impact on her life. She now understood more-than-ever the kind of impact that the domestic abuse had on her life.

On another level, she confirmed that on a positive note, our interview helped her take a stand with someone who was emotionally abusing her in the context of an intimate relationship. Sandra wrote that in a relationship, she does her utmost not to attack even when the other person does her or his best to provoke her. She tries to keep as calm as possible because she believes that retaliation is futile and would make her look foolish. Thinking about this after our interview, Sandra realized that her reaction is probably “…an innate reaction to the violence that I have lived as a child”. “When I was a child, I never understood the screaming, shouting and hitting… and I remember thinking and saying to myself that when I will grow older, I don't want to be like that”. But she also realized that she needed to balance things out and not allow others to abuse her just because she is not a bully. She learned that she must be respected for who she is, and to acknowledge and label as early as possible any form of abuse towards her. When contacting Sandra's psychologist as part of the continuity of care after the interview, the psychologist also told me how our interview had been a catalyst for Sandra in their work together and they were now working on some of these same insights that had been set off by the interview.

In another example, Carina's interview was also another encounter that proved meaningful for her and for me as well. I came out of the interview feeling a renewed burst of energy for life. On my way home in the car, I thought to myself “What an interesting phenomenon–that both of us had felt energized by our encounter!” To me, Carina's way of living her life was a metaphor for resilience, broadly defined as “the capacity to rebound from adversity strengthened and more resourceful” (Walsh, 2006, p. 4). She was a powerful example of fighting life's hard deals, and her hunger to have her life back impressed me enormously.

Her story was very tragic and full of experiences of suffering. Yet, these stories were also complemented by insights about what she had learned from these experiences and by statements of what inspires her to move forward in life. As I write about her experiences, I recall the concept of vicarious resilience (Engstrom, Hernandez, & Gangsei, 2008). These researchers empirically studied how therapists working with traumatized clients learn about coping with adversity from them and how these clients' narratives can have a positive effect on the therapists. Taking Carina's point of view, I also wondered about her insights, and how her way of trying to make sense of what she went through and what she had learned from her experiences can be seen as way of healing from traumatic experiences and as a form of making meaning from loss (Neimeyer, 2001).

On another level, as an interviewer/clinician, I found myself easily warming up to her. Perhaps in a way, she represented my ideal client–very much in touch with her feelings and her suffering and yet full of insight and determination. However, I was also very conscious of the fact that because we were meeting for just one interview, even though I wanted to ask her more about various other experiences, I felt I owed her the privilege of actively listening to her story without my interruptions. At the same time, there were moments where I felt I needed to give her feedback about her strengths and resilience. I also felt the need to share the insights that I had learnt about relationships through life experiences, through training, and through the interactions with other clients. I was fully aware that insights need to be learnt by living through life experiences (Rogers, 1942) and, not just listened to. At the same time, I wanted to give her feedback, to give her something back as a thank you for what she had given me through the sharing of her experiences.

Following the interview, Carina told me that despite going through a difficult time during the interview, she had re-connected with her sense of resilience and with other challenges in the past that she had overcome in her life. This was also re-iterated by her psychologist when I spoke to her and asked her about how Carina was after being interviewed and whether she had experienced any undue distress.

Conclusion

In this article, we have discussed some of the ethical dilemmas for clinicians conducting qualitative research about sensitive topics. We have shared our examples from practice, recognizing that these dilemmas may also be relevant to other researchers from the field of nursing, medicine, and also psychotherapy. At the same time, we are aware that the feedback from our participants was limited and anecdotal. In future research, it would not only be interesting to obtain systematic feedback from participants but also to triangulate this with the feedback of the interviewers.

We now conclude with some recommendations of good practice that are based on our experience in research and our reading of the relevant literature. We do not in any way assume to have come up with prescriptive guidelines. Our hope is that this paper will trigger debate amongst qualitative researchers and we welcome the feedback of other researchers/practitioners in the field.

Having access to clinical training or clinical supervision in our experience was an important resource from which to draw from, to act ethically in the field and to disentangle some of the complexity of the different levels of interaction between researcher and participant. It was also important to help clarify the impact of the role of the clinician and that of the researcher. The use of the reflective diary or journal and detailed transcripts of the interviews were also useful supervision tools to ensure ethical considerations and dilemmas were addressed in the research process.

Our supervision discussions also helped us appreciate the importance of having one's supportive networks when doing research with potentially vulnerable populations about sensitive topics, especially when these discussions helped further the understanding of distressing effects on the participants and on the researchers. Our experience showed that it was useful to leave enough space between one interview and another to process any information that may be distressing, and not to schedule appointments when one is fatigued, for example after a long day at work.

In terms of recruitment of participants, we have found it useful to work collaboratively with other professionals, especially the psychologists working with the participants, to ensure the well-being of the participants. As a researcher, these professionals were the safety nets for participants and we were reassured that a one-time research interview would be reflected upon and deconstructed, if needed, in the context of a safe, therapeutic relationship. These professionals were also available to support the participants if further emotional help was required.

It was also important to give participants various opportunities to give their consent during the different stages of the research and not just the initial stages of the research. Although in this study this continual consent was obtained verbally, in hindsight, it would have been preferable to outline this process in the information sheet or in the consent form itself.

It was also essential that debriefing happen after each interview, even if the participants did not demonstrate any undue distress. It was also a way of getting feedback from participants about the participants' comfort level during the interviews. This feedback was an indicator of how the research interview was going and can be a way of ensuring “ethics in practice.”