Conducting Sensitive Interviews: A Review of Reflections

• 1 Introduction

  There is no agreed definition of sensitive research (Dickson-Swift et al. 2007, 2009), although the term most often refers to research about emotionally difficult topics that require participants to face issues that are deeply personal and possibly distressing (Cowles 1998, Johnson & Macleod Clarke 2003, Lee 1993:109). Sensitive research can also involve topics that may be taboo, intrusive, stigmatising (Demi & Warren 1995), illegal and potentially dangerous (Carmack & DeGroot 2013:326, Dickson-Swift 2009, Lee 1993:4). Conducting sensitive research can present a number of serious challenges. Participants may find it distressing to discuss painful and embarrassing issues (Demi & Warren 1995, Dickson-Swift et al. 2008, Durham 2002, Johnson & Macleod Clark 2003, Owens 2006), and unless research is conducted in a sensitive manner, participants may be re-victimised, for example by re-experiencing a trauma (Lee & Renzetti 1990).
  Sensitive interviews can also take an emotional toll on researchers. Researchers have reported feeling distressed as they listen, often repeatedly, to people’s experiences of trauma (Bloor et al. 2010, Booth & Booth 1994, Corbin & Morse 2003, Johnson & Macleod Clarke 2003, Lee & Renzetti 1990, Sherry 2013). Researchers have also experienced insomnia and nightmares, feeling exhausted, anxious, stressed and even depressed (Cowles 1999:173). Some have found that listening to multiple similar stories have left them feeling de-sensitised. Others report feeling a heightened sense of their own mortality and vulnerability (Dickson-Swift et al. 2006, 2007). As Booth and Booth (1994:422) note, researchers can become stressed because of the “strain of witnessing and sharing the anguish of the informant, and the strain of coping with the feelings they release in oneself.” Watts (2008:5) explains that building rapport requires researchers to seek out comparable experiences, and it is this that produces emotional reactions. As she states in relation to conducting interviews with cancer patients:
  

  Specifically, when working with participants I look to my inner self and my fears, searching for comparability of experience. Questions of shared feelings nag at me because I am troubled about my own relationship to what I want to learn. This concern is difficult to articulate but is rooted in what can best be termed as an unspoken hierarchy of ‘authentic’ fear with my anxiety about being a potential cancer patient in stark contrast to the reality of the daily lived fear of participants. This negative emotion, that I feel unable to voice within the research setting, has prompted much reflection on the issues of loss and suffering, particularly the nature and experience of suffering that… occurs when a person is confronted by their impending death (Watts 2008:5).

  
  In addition, other members of a research team are also vulnerable. Transcribers are required to capture the emotional nuances of interviews, and this involves listening intently and repeatedly to traumatic data (Bahn & Wetherill 2013:28, Gilbert 2001a, Malacrida 2007). The need to pay close attention to the data also leaves coders vulnerable to distress (Malacrida 2007, Woodby et al. 2011).
  Despite these risks, however, researchers often lack advice when conducting sensitive research (Bloor et al. 2010, Hubbard et al. 2001, Johnson & Macleod Clark 2003). For the most part, researchers report their methods in terms of the technical details, such as recruitment methods, response rates and techniques for data collection and analysis. Some socio-legal researchers have provided invaluable advice on dealing with these issues (e.g. Banaker & Travers 2005, Halliday & Schmidt 2009). However, discussion about other important methodological issues, especially those that require a deeper level of critical self-reflection, are less forthcoming.
  Some researchers have provided reflections on how they have conducted sensitive research, including studies that have asked respondents about their experiences of research (Campbell et al. 2009, Goodrum & Keys 2007). Dickson-Swift et al. (2007, 2008, 2009) have also interviewed researchers who specialise in this type of research. This article provides a review of these reflections, focusing specifically on the conduct of sensitive interviews. Often, sensitive interviews are conducted by researchers with relatively little experience, such as graduate students or casual research staff, and researchers often work in isolation (Gilbert 2001a, Johnson & Macleod Clarke 2003). In addition, legal researchers often do not have the same level of training in research methods as researchers from social science disciplines (Genn et al. 2006). This leaves empirical legal researchers especially prone to carrying the emotional burden of their work without adequate support. We hope that this review will provide a useful resource for empirical legal researchers wanting to conduct sensitive interviews.
  This article is in four parts. First, we discuss the need for researchers to follow ethical guidelines in order to reduce potential harm to respondents, focusing specifically on the need to obtain informed consent and maintain confidentiality. Many researchers assume that the presence of ethical guidelines will ensure that researchers adequately deal with emotional issues and mitigate harm; however, sensitive research often involves challenges not covered by the guidelines (Alty & Rodham 1998, Bahn & Weatherill 2013). The second part of this article looks at these additional issues, including providing emotional care of participants before, during and after the interview. Most reflections focus on the impact of sensitive research on participants, whereas researchers also need to ensure their own emotional care (Hubbard et al. 2001). The third section looks at some of the ways in which researchers can do this, as well as arguing that institutions and Institutional Review Boards should also pay attention to researchers’ emotional well-being. Finally, an example of conducting sensitive interviews is provided on the basis of the author’s experiences in relation to conducting interviews with clinical negligence claimants.

• 2 Protecting Respondents through Ethical Guidelines

  The main way in which risks are assessed and mitigated is through the use of ethical guidelines. The development of ethical guidelines has followed people being seriously harmed by researchers (Faden et al. 1986, Israel & Hay 2006). The Nuremberg Code (1947), which provides the basis for most social science ethical guidelines, was developed following the Nuremberg Military Tribunal, which revealed the horrors of Nazi medical experiments. The Nuremberg Code sets out ten standards that must be met by researchers:
  

  1. Research participants must give their voluntary consent.

  2. Research needs to be beneficial to society.

  3. Research must be appropriately designed.

  4. Researchers must avoid inflicting harm.

  5. A study cannot commence if there is a risk of causing death or disability.

  6. Risks to participants should not exceed benefits.

  7. Adequate preparations must be made to protect against harm.

  8. Research must be conducted by qualified researchers.

  9. Participants can withdraw.

  10. Research must cease if there is cause to believe that its continuation will cause harm.

  
  Some jurisdictions (e.g. the United States, Canada, Australia, Norway) have incorporated these principles into laws. In these jurisdictions, all empirical research that may cause harm must be reviewed before the commencement of the study, and reviews are largely conducted by Institutional Review Boards (IRBs). IRBs usually require researchers to ensure that the benefit of participating in research outweighs the risks of harm. Harm means physical as well as psychological, social and economic damage. This principle does not prohibit the risk of harm; however, it does require that the potential for harm is outweighed by the potential for benefit. If there is a risk, researchers should first attempt to minimise it, and second, they should ensure that potential participants are fully informed of this risk as part of the process for seeking consent.
  Some researchers have welcomed the formalisation of ethical guidelines. For instance, Israel and Hay (2006:2) explain that ethical guidelines promote trust, improve rigour and research integrity, help ensure that research is beneficial, and “offers the potential to increase the sum of good in the world.” They also acknowledge that there have been incidents where researchers have not adhered to ethical standards, resulting in harm to individuals and communities.
  Not all researchers have taken this view, and IRBs have been criticised for being meddlesome, overly interventionalist and highly risk adverse (Bloor et al. 2010:50). A survey of researchers investigating the needs of those bereaved by suicide found that most felt that IRBs play an important role in ensuring that research adheres to ethical guidelines. However, researchers also believe that IRBs often impose restrictions without sound knowledge of the risks posed by the research (Moore et al. 2013). Researchers have also stated that the way in which IRBs perceive risks is highly variable and inconsistent (Larson et al. 2004). Different IRBs have assessed the same project as presenting both a minimal and a high level of risk (Hirshon et al. 2002, McWilliams et al. 2003).

• 3 Obtaining Informed Consent

  The first principle laid out in the Nuremberg Code is that researchers must obtain informed consent from potential research participants in advance of data collection. Obtaining informed consent involves disclosing to potential participants the research aims, what participation will involve, how the research is to be used, how confidentiality will be maintained and limits to confidentiality and any risks that may arise from participation. Participants need to be able to comprehend this disclosure, consent must be voluntary, they must be competent to provide consent, and consent must be established, usually by signing a consent form (Booth & Booth 1994:417, Faden et al. 1986:27).
  Simply being provided with information does not ensure that a potential participant is informed. Some guidelines require, wherever possible, that prospective participants are provided with written information concerning the study. Written information allows prospective participants to read the document in their own time. However, information sheets are often overly long, poorly formatted and written in a style that many people find difficult to understand (Jefford & Moore 2008). Some concepts are especially difficult to convey, such as randomisation (Criscione et al. 2003, Featherstone & Donovan 1998), and legal language can also be confusing (Dawson & Kass 2005:1215). These problems may be exacerbated by IRBs; for instance Paasche-Orlow et al. (2003) show that IRBs commonly require researchers to include text that falls short of their own readability standards.
  Some ethical guidelines encourage researchers to talk to prospective participants face-to-face, which can be especially useful for delivering information to people with poor literacy skills or people who feel intimidated by written documents. Some researchers have also found it useful to use community meetings or other forms of group discussions to explain their research and allow people to ask questions (Dawson & Kass 2005:1215). Most researchers conducting sensitive interviews do not appear to rely solely on written information sheets; rather, they use a mix of methods to communicate the details of their study.
  Difficulties in obtaining informed consent have led to further discussion about whether it is ethical to include vulnerable participants in research at all (Darra 2008). Vulnerable respondents, such as children, prisoners, people with disabilities and the incompetent elderly have not always been provided with adequate protection from harms caused by researchers (Horner 1999). Some researchers see obtaining informed consent as a ritual involving securing a signature on a form that the participant has not read or understood, after which they can do as they like (Faden et al. 1986, Jacob 2007, Murphy & Dingwall 2001). Many research participants also do not fully understand what they are consenting to (Cohn & Larson 2007), and people who do not understand are more likely to agree to participate (Dawson & Kass 2005:1220). Demi and Warren (1995:192) also warn that vulnerable participants may agree to take part in an interview for inappropriate reasons. These include deference to an authority figure or the need to talk to someone. Vulnerable participants who are repeatedly exposed to bureaucratic treatment may sign the consent form without reading it as they see it as another bureaucratic form (Wong 1998:187). Vulnerable participants may not feel able to withdraw if they become uncomfortable (Liamputtong 2007), or may struggle to set limits on what information is revealed (Glazer 1980).
  In reply, it has been argued that research protocols set by IRBs may not be perfect, but that they generally provide good enough protection to research participants (Israel & Hay 2006). Liamputtong (2007) questions whether research with vulnerable participants can ever be fully ethical, but she also acknowledges that to exclude vulnerable respondents from research altogether could lead to further disempowerment. Therefore, it is important that researchers are attuned to ensuring that consent is informed and freely given, and understand that this requires more than obtaining a signature on a consent form.
  A number of strategies can be used to assist in obtaining informed consent. Some researchers advise that contacting potential participants before sending out an information sheet helps with understanding (Stevens et al. 2009:498-500). Information should explain the possibility of emotional responses such as distress (Brzuzy et al. 1997:80, Cowles 1988:167), and should be written in an emotionally sensitive language (Goodrum & Keys 2007). For instance, Brzuzy et al. (1997:80) suggest that participants should be informed about possible anticipatory reactions, such as anxiety, reduced concentration, flashbacks and distress. They also suggest that providing information about potential consequences of participation may help participants to normalise their feelings in advance of their occurrence. Some researchers have used assessments of comprehension to ensure that people have understood the information that they have been given (Wirshing et al. 1998). Researchers should also be careful to avoid inferring that initial expressions of interest and questions constitute consent (Stevens et al. 2009).
  Further, Durham (2002) suggests that participants should be informed of the researcher’s ethics, values, knowledge and experience. For instance, Durham (2002) suggests being explicit about whether you are committed to pro-feminist or anti-oppressive values, the nature of your previous research projects and what interviewees may gain from participating. This depth of information allows respondents to be aware of whether their participation will be beneficial or whether it only suits the researcher’s aims. Knowing that the researcher is skilled also helps respondents feel that the interview will be conducted in a way that is sensitive, informed and non-judgmental.
  Researchers should also be aware of the limitations of relying on informed consent given at the onset of a study. Bahn and Wetherill (2013:29) advise that researchers need to be prepared to make decisions ‘on the run’, and that sometimes an interview needs to be terminated in order to prevent harm. Researchers should be aware of the pressures to continue even though it is clear that the risks outweigh the benefits. These may include pressure from institutions for research outputs, PhD timelines (Owens 2006) or the researchers’ own ambitions (Durham 2002:427).
  Participants may not be the only people that need to provide consent. Parents or carers of minors may need to provide consent, and often young children are asked to provide assent rather than consent (Morrow & Richards 1996). Researchers may also need to obtain consent, or at least access, from institutions such as schools (Barker & Weller 2003) and prisons (Trulson et al. 2004). This can create gatekeepers who may deny access to participants in order to protect them from distress (Taylor 2002:146). For instance, in a study of parents of children with life-threatening illnesses, Stevens et al. (2009:499) found that clinical staff were sometimes reluctant to inform families of the study as they were concerned that participation would cause distress. Gatekeeping can cause other problems. Stevens et al. (2009:499) also found that staff were unable to provide precise information about how many potential participants were excluded from the study, how many were contacted but then declined to participate and their characteristics. (We discuss this problem further in relation to interviewing clinical negligence claimants.) Different staff members also held different views concerning vulnerability, which added further unreliability.
  Sensitive research also raises an additional question of the best time to contact prospective participants. For instance, Cowles (1988) interviewed people one month after the sudden violent death of a loved one. Some participants reflected that they could not have coped with an earlier interview. Research conducted too soon can leave participants feeling emotionally overwhelmed, and researchers may be seen to be too intrusive (Cowles 1988:164). In contrast, Bentley and O’Conner (2014) conducted interviews five months after the death of a family member, and the majority of participants reported that they could have handled the interview earlier and that they would have welcomed an opportunity to discuss their feelings. In this instance, participants wanted to be allowed to decide for themselves when they should be involved in an interview.

• 4 Maintaining Confidentiality

  The other ethical requirement that is usually regulated by IRBs is confidentiality. Researchers hold a relationship of trust with respondents, and it is generally considered unethical to breach this trust. Ethical guidelines usually require that participants are informed about how the information they provide will be used, data is to be stored and results will be presented. To ensure that a participant’s right to privacy is protected, researchers usually anonymise their data. This has generally been taken to mean that identifiers are removed from data. Some researchers also employ pseudonyms or numerical identifiers to protect a participant’s identity. These measures provide some degree of anonymity, and many researchers accept them as being sufficient (Beyleveld & Townend 2004). However, confidentiality cannot always be assured when conducting sensitive interviews.
  Sensitive research can reveal instances where the participant is at risk of harm (Lee & Renzetti 1990), involved in illegal activities (Kvale 1996:115), or have threatened harm to someone else (Cowburn 2005). For instance, Findholt and Robrecht (2002) faced instances of young people disclosing that they had committed statutory rape. Similarly, in interviews with convicted sex offenders, Cowburn (2005) was given previously unknown information relating to someone being at risk of serious harm. Researchers may have a moral or legal duty to disclose such information.
  The need to disclose raises numerous challenges for researchers conducting sensitive interviews. In a study of how researchers negotiate disclosure, Wiles et al. (2008) found that researchers will report instances of harm, but often not of illegal activities. Similarly, Israel (2004) reports that to protect participants, researchers have lied to police, prosecutors and correctional authorities. Some researchers have also accepted the consequences of failing to disclose. For instance, Rik Scarce was imprisoned for 159 days after refusing to provide a US federal grand jury with access to confidential interviews (Scarce 1994).
  The potential need to disclose highlights that researchers must plan ahead, as demonstrated by Abel et al.’s (1987) study of the behaviour of sexual offenders. In this US-based study, Abel et al. (1987) asked convicted sexual offenders to disclose unreported sexual crimes. The researchers attempted to maintain confidentiality by asking only for general features of offences, de-identified data by allocating each participant a unique code, and information that linked the codes to identifying material was then stored outside of the US to prevent data being subpoenaed. The ethics of Abel et al.’s (1987) work has been defended as it resulted in a significant shift in thinking about how offenders target victims that would not have been possible without confidentiality being guaranteed (Cowburn 2005:53).
  Abel et al.’s (1987) study, however, involved a survey rather than in-depth interviews, which pose a higher risk of perpetrators and victims being identified, and thus creates a strong imperative to disclose (Cowburn 2005:54). Cowburn (2005:58) suggests that the limits of confidentiality need to be made clear when seeking consent, and participants should also be reminded regularly during the interview process. Cowburn (2005:60) also suggests that the decision to breach confidentiality needs to consider the nature of the harm (whether the harm is already known/unknown to authorities, whether the harm is general/specific), identity of the perpetrator (whether information is first/second hand) and the identity of the victim (whether the threat is general/specific). Some researchers have argued that past harms are of no concern (Scully 1990); however, Cowburn (2005:60) contends that both previous and future threats should be addressed.
  Dealing with disclosure can also be emotionally difficult for researchers. Many researchers feel uncomfortable having to decide whether to breach their relationship of trust or to leave someone at risk of harm in order to protect a participant’s identity. Booth and Booth (1994:423) suggest that in these circumstances, the “interviewer should not be expected to carry the moral burden of their knowledge alone.” They recommend using a reference group of experienced outsiders with whom the researcher can discuss anonymised material that poses ethical dilemmas. Similarly, Demi and Warren (1995:198) suggest that formal support should be made available, including regular meetings to address ethical dilemmas and opportunities for immediate consultation with experienced colleagues.
  The nature of data obtained from in-depth interviews can also limit confidentiality. Damianakis and Woodford (2012) warn that changing key characteristics of participants, including the use of pseudonyms, is not always sufficient to guarantee anonymity. The ability to identify participants from quotes and other detailed informed is especially difficult when participants come from a closed community, share professional or personal networks, or when the sample population is small. Vivid descriptions of other people within the community can unintentionally reveal a participant’s identity, as well as those of other people in the community or other participants. Data such as the names of friends, locations, employment, recreational pursuits and specific events can all identify a participant. Damianakis and Woodford (2012) illustrate the risk of identification in terms of public policy officials who may be identified if they are critical of government policies or push specific agendas, especially if they are known by their colleagues to hold these views.
  A guarantee of confidentiality can result in participants sharing sensitive information that they would not usually reveal and without fully realising how this material may be used (Dickson-Swift et al. 2007, Saunders et al. 2015, Wong 1998:193). For example, when interviewing families of patients with a severe brain injury, Saunders et al. (2015) noted that some family members stated that they wished their relative had died instead of being severely incapacitated. Revealing the specific details of the relative’s brain injury could have then linked participants with these comments.
  The need to protect participants’ identities can leave researchers with the difficult decision of omitting crucial interview data from publications (Wiles et al. 2008), which is especially problematic when the participant’s entire narrative is needed to contextualise the research (Saunders et al. 2015). Anonymising interviews can also be time consuming and costly (Damianakis & Woodford 2012, Saunders et al. 2015).
  It is not just the final publications that may potentially identify participants. In interviews concerning the impact of parental substance abuse on children, Demi and Warren (1995:197) sent information to participants in plain envelopes, no information was included in correspondence that could have revealed the participants’ stigmatising status, and telephone calls were made so as not to identify the research aims. The researchers also kept a low profile in the community, and when pressured to identify participants pointed out that the study involved a comparative group, and that participants could have been members of the control group. Similarly, in a study involving women from rural communities who use crack cocaine, Brown (2003) used several strategies to conceal that nature of the research from members of the community. Researchers posed as advisors for an education outreach service that provided information on cancer, cardiovascular disease and HIV prevention. They conducted some interviews within the service’s office, took information from the service into participants’ homes, wore T-shirts and badges similar to those of advisors, made outreach visits to other residents in the local area, and some participants became clients of the outreach service.

• 5 Conducting the Interview

  It is important for researchers intending to conduct sensitive interviews to be adequately prepared. Researchers need to be trained (Dickson et al. 2007), which may include practical exercises to develop interviewing skills, such as watching model interviews and participating in mock interviews (Campbell et al. 2009). Research participants have stressed that researchers need to have prior knowledge about the topic. This may include understanding the academic literature, although research participants have also stressed that academic expertise does not necessarily attune a researcher to the type of personal experiences that may be raised in an interview (Campbell et al. 2009). Brzuzy et al. (1997:80) advises that researchers should prepare for worse-case scenarios, where the participant may become distressed or raise ethical dilemmas. Stevens et al. (2009:499) also suggest that preparation could involve researching the participant’s backgrounds before making contact in order to better understand the participant’s individual circumstances.
  Preparation should also focus on ensuring that interview topics are appropriately sensitive. For instance, Goodrum and Keys (2007) found pilot interviews useful for highlighting terms, such as ‘closure’, which interviewees thought should be avoided. Participants also appreciate being asked questions that focus on their positive experiences. This could include questions on participants’ ability to cope, their hopes and ability to recover and survive in the midst of adversity, rather than their deficiencies, weaknesses and failures (Sanford 1990, Turnball et al. 1988). Participants’ reactions to questions will be highly individualised, and researchers should avoid assuming what topics may cause distress (Campbell et al. 2009, Goodrum & Keys 2007). Topics routinely asked in questionnaires and assumed to be emotionally safe, such as the participant’s current income or educational attainment, may also evoke emotional responses when further explored within an interview (Turnball et al. 1988).
  The first step in conducting the actual interview is to establish rapport, which refers to the relationship of trust between researcher and participant. Rapport can be used to gain access to a research setting, motivate a participant to agree to an interview (Agar 1996), negotiate uneven power relations (Marx 2001) and open up conversational space (Owens 2006). Researchers often try to build rapport by starting an interview with general questions (Goodrum & Keys 2007, Stevens et al. 2009:489). However, some respondents have a pressing need to discuss emotional issues immediately, and it is only when they have offloaded that the rest of the interview can proceed (Stevens et al. 2009:498). For instance, Booth and Booth (1994) explain that they had initially planned to use the first meeting to introduce their study rather than for data gathering. However, it was clear that some people wanted to talk straight away. The researchers took their cue from the respondents, and adapted their research design according to the participant’s emotional needs.
  It is likely that participants will experience strong emotions during the interview, and researchers need to be prepared for participants to cry, become angry, struggle to regain control and abruptly change topics (Cowles 1988:167). In response, researchers may need to fall back on a ‘bailout’ protocol, such as moving to factual questions, taking a break or offering to resume at another time (Cowles 1988:168, Hallet 1995:503, Turnball et al. 1988). However, it is also important that researchers do not evade difficult topics. Avoiding topics will close down conversational space, and so it is preferable to take a direct approach and acknowledge the participant’s experiences, further explore their feelings, and ask why they are upset or uncomfortable. A direct approach indicates that it is acceptable, rather than shameful, to discuss painful experiences (Cowles 1988:171, Owens 2006:1176).
  Participants who are very distressed may also need assistance. There is debate within the literature concerning whether researchers should intervene. This is especially an issue when the research has a dual professional identity, for instance the researcher may also be a counsellor (Hoeyer et al. 2005). Some authors advocate a hands-off policy as intervening could bias responses (e.g. Bromet & Havenaar 2006). In contrast, others argue that researchers are ethically obliged to support participants to debrief (Alty & Rodham 1998:180), and offer intervention in order to minimise harm (Mertens & Ginsberg 2008). A possible middle ground may be to intervene when requested or in an emergency (Booth & Booth 1994, Cowles 1988:168, Turnball et al. 1988). Most researchers suggest providing a list of referrals to support professionals, including crisis contact number, rather than providing intervention themselves (e.g. Beale et al. 2004, Brzuzy et al. 1997:80, Stevens et al. 2009:49).
  It is important that sensitive interviews are not rushed, and that researchers allow time to develop rapport, participants are able to cry, take a break or move between topics as need be. Participants also need to feel that they can express their feelings without being constrained by time (Cowles 1988:165-166, Stevens et al. 2009:498). Cowles (1988:167) also warns that scheduling interviews late in the day may leave researchers and participants alike vulnerable to insomnia and nightmares.
  Finally, Booth and Booth (1994:417, 422) explain that researchers need to consider how to close the interview. The researcher needs to gauge from the participant the moment when the interview can be drawn to a natural halt, and be wary that to continue may leave the participant exhausted or turn the interview into an interrogation. The task of completing a research project can also be difficult, especially if participants have a restricted social network and the research is one of the few opportunities participants have to discuss their feelings. Researchers have also expressed discomfort at withdrawing knowing that the participant is facing a difficult time, for instance when the participant has deteriorating health or is dying (Burr 1996:172, Rosenblatt 1995). Booth and Booth (1994:417) advise leaving participants to determine the pace at which researchers withdraw, and this may mean that the relationship between research and participant endures beyond the study. Researchers also suggest that it is important to follow up so that participants are offered the opportunity to discuss emotional side effects, and are provided with further support if necessary (Brzuzy et al. 1997:80).

• 6 Should Researchers Reciprocate?

  The ethical issues faced by sensitive researchers go deeper than simply being prepared for the interview and asking questions in an emotionally sensitive manner. For most researchers, the conduct of sensitive interviews requires self-reflection concerning the relationship between themselves and participants, especially if this relationship is unequal. One important question that researchers need to consider is whether they should ‘give something back’ to participants.
  The level of rapport needed to conduct sensitive interviews often leads to researchers feeling as if they should reciprocate (Campbell 2002, Corbin & Morse 2003, Dickson et al. 2008, Johnson & Macleod Clark 2003, Sherry 2013). This sense of wanting to reciprocate can also endure after the project is completed, as explained by Bahn and Wetherill (2013:33):
  

  While I was working on something totally different Mr and Mrs < > kept popping into my head. I felt like visiting them, you know, just to pop in for a cuppa and do something; something that would make things better for them.

  
  The research literature usually discusses reciprocity in relation to incentives. Researchers are warned that using incentives to overcome reluctance to participate is unethical as it is contrary to the principle of voluntary consent. Incentives can also bias results, as participants may feel that they should say what the researchers want to hear (Rice & Broome 2004:168-169). An example of how incentives can alter behaviour is provided by Wong (1998:191), who paid participants a financial stipend. Wong (1998) then faced people who wanted to participant only because of the money, and consequently interviews were monosyllabic and inarticulate. Wong (1998:191) also reported that she felt betrayed when she realised that participants were not really interested in her study.
  In contrast, some researchers support the use of incentives, although they do not necessarily need to reward participation. Booth and Booth (1994:417) explain that there are two main models for conceptualising rapport. First, rapport can be seen as a form of mutual exchange, with trust being established through reciprocal relations where both researcher and participant give each other something they desire or need. This approach is demonstrated by Demi and Warren (1995), who support the idea of providing something that participants want and value in order to secure an interview. Participants may value money or gifts, but incentives could also include the provision of childcare and transport in order to facilitate interviews, access to services and information, and the opportunity to tell their story.
  Second, rapport can be seen as a form of moral currency, where trust is dependent on the researcher’s actions and attitudes. In this model, reciprocity demonstrates that participants are valued in their own right, and not just for their consent to participate in an interview. This approach is demonstrated by Taylor (2002:156), who argues that sensitive research should involve a sustained commitment to participants. This may involve spending extended time within participants’ communities and showing willingness to share knowledge, such as producing a newsletter to present findings, placing articles in local newspapers or giving radio interviews.
  To sustain trust it is important that any promises to reciprocate are fulfilled. Respect is shown through keeping appointments, returning telephone calls and following up on actions that may be promised such as obtaining information. Trust will also be diminished if researchers overinflate the benefits that may be provided by research (Booth and Booth 1994:418). Some participants will see that benefits from participation will be direct and immediate, when it is often more likely for benefits to be indirect and provided to other members of society rather than the participant (Rosenthal 1994). Booth and Booth (1994:418) suggest that researchers should clearly explain their role, and avoid being seen as a benefactor who may not be able to deliver. Similarly, Demi and Warren (1995:198) advise that researchers should be upfront about problems that researchers cannot fix.

• 7 Emotional Care for Researchers

  It is clear that researchers, and not just participants, need emotional care. This awareness needs to inform the preparatory stages of research. Preparation may involve having discussions beforehand with other researchers and support professionals, such as counsellors, especially if the researcher is relatively inexperienced (Cowles 1988:174). Prior work may not fully prepare a researcher, but is better than nothing. For instance, Cowles (1988:174) reported that she was not fully prepared for either the quantity or the intensity of emotions expressed by participants; however, she had at least partially anticipated that participants could become upset, and so she was not alarmed at her own emotional responses.
  Cowles (1988:174) reports that when she sometimes cries alongside respondents, they seem surprised. She attempts to explain that her emotions should not cause embarrassment, and hopes that her response is interpreted as an expression of empathy. Other researchers have also seen that their emotional responses are an indicator of rapport, and have argued that expressing emotions is an important part of the research process that should not be avoided (e.g. Gilbert 2001b). However, some research participants have stated that they would prefer to have the researcher listen with compassion and patience rather than overt distress (Campbell et al. 2009). Participants have reacted to researchers’ distress by changing the topic or feeling as though the researchers have stopped listening in order to deal with their own feelings (Goodrum & Keys 2007).
  The main form of support for researchers suggested by reflections is debriefing. This provides an opportunity for researchers to express emotions, make sense of what may initially seem to be inexplicable and relate the participants’ narratives to their own lives (Beale et al. 2004, Johnson & Macleod Clark 2003). Debriefing may also alleviate feelings of isolation, loneliness and being de-sensitised (Sherry 2013), and knowing that anxiety and distress are common reactions can be reassuring, especially to inexperienced researchers (Cowles 1988:173).
  Debriefing can take many forms. For instance, keeping a journal allows a researcher to express his or her emotions and provides an opportunity for reflection. However, keeping a journal is typically a solitary practice (Carmack & DeGroot 2013:326) and is therefore unlikely to help with feelings of isolation. Keeping a journal is also, typically, an informal practice, put in place at the researcher’s own initiative. Indeed, for the most part, it appears that emotional support for a researcher is informal and ad hoc.
  Formal systems could include scheduling regular meetings to review the research, discuss transcripts and coding, share interview techniques and emotionally debrief (Campbell et al. 2009). Formal support systems also ensure that senior researchers take responsibility for providing support, rather than leaving junior researchers to ask for assistance (Bahn & Wetherill 2013:28, Brzuzy et al. 1997:81, Bloor et al. 2010, Darra 2008). The formalisation of support would require building time for debriefing into schedules and budgets (Bahn & Wetherill 2013:28, Cowles 1988:166).
  Team meetings are sometimes used for debriefing; however, it should not be assumed that meetings provide researchers with appropriate support (Stevens et al. 2009:504). In teams where only one individual is responsible for interviews, team meetings do not allow for reciprocal sharing of emotions and experiences. Researchers may need external support in order to have access to someone with direct experience of conducting sensitive interviews, although this can raise issues of confidentiality. In addition, Cowles (1988:175) recommends that more than one colleague should be available for debriefing, as it is helpful to share the emotional burden and to ensure that someone is always available (Cowles 1988:175). Similarly, Beale et al. (2004) recommends building egalitarian working models so that senior researchers are readily available, and so that all members of the research team, including transcribers, are included.
  The formalisation of support systems also requires oversight. Bloor et al. (2010) show that researchers rarely seek support from university human resource departments or counselling services and do not undertake risk assessments, and yet these services are readily available. Nor is the failure to provide formal support a symptom of the failure of the management of risk in society generally, as risk assessments, safety training and debriefing are commonplace in other sectors. Instead, risk management does not appear to be part of the university culture (Bloor et al. 2010:51). Universities have a duty of care towards their staff, but grant holders, project managers, PhD supervisors and heads of departments are often remote from the everyday practice of research (Bloor et al. 2010, Hubbard et al. 2001).
  It has been suggested that IRBs should accept formal responsibility for care towards researchers conducting sensitive research. Additional measures may also be possible, such as funders asking peer reviewers to comment on researcher safety and care, including issues in research methods courses, in-house training for PhD supervisors and grant holders, and periodic health and safety audits that include the examination of provisions for researcher safety (Bloor et al. 2010:50-51; Bahn & Weatherill 2012:26).

• 8 Researching the Experiences of Clinical Negligence Claimants

  Our review has covered the advice given by researchers who have conducted sensitive interviews. Most of these researchers, however, come from the health sciences and thus have not focused on socio-legal topics. The final section of this article provides insights into the conduct of sensitive interviewing from the direct experiences of one of the authors, who has been involved in two projects aimed at understanding the experiences of clinical negligence claimants. The studies focus on why claimants sued, their experiences of making a claim, their relationship with their solicitor and claimant satisfaction with claim outcomes (see Stephen et al. 2012, Melville et al. 2014).
  Most previous research on clinical negligence claiming has analysed medical records (e.g. Brennan et al. 1991, Hiatt et al. 1989), insurance company records (Brennan et al. 1996; Danzon 1984; Studdert et al. 2006), regulator databases (Black et al. 2005, Vidmar et al. 2004) and jury verdicts (Bovbjerg & Bartow 2003:3) to either examine the outcomes of claims or predict the incidence of adverse events. Studies have also drawn on claimants’ medical files to predict the incidence of adverse events and analyse circumstances that activate claims. This focus on outcomes means that little is known about the process of making a claim. There is also a large body of research involving medical professionals rather than claimants. Research has examined medical professionals’ experiences of being sued, including doctors’ understandings of what constitutes and causes medical errors (Blendon et al. 2002, Jagsi et al. 2005), and the adverse impact of litigation on medical professionals, including high levels of stress (Charles et al. 1985, Nash et al. 2007). This focus means that our understanding of clinical negligence claiming is potentially one-sided.
  Previous research has generally overlooked the perspective of claimants themselves. The few studies that have been done have predominantly used quantitative methods to look at claimants’ motivation (Hickson et al. 1992, Huyke and Huycke 1994, Shapiro et al. 1989, Sloan & Hsieh 1995, Vincent et al. 1994), with the exemption of Relis (2006), and have left important questions unanswered. For instance, they do not investigate how claimants view the incident that led them to sue, their experiences of making a complaint or their relationship with medical or legal professionals. It was also decided to avoid a quantitative approach as it was suspected that claimants may have sued as they felt that their problems had not been listened to, and that it was important to use a research method that allowed claimants to give a full account of their perspective from their own point of view.
  The first study involved 30 in-depth interviews with people who intimated a clinical negligence claim from January 2006 to June 2009. The sample consisted of 18 women and 12 men. Five claimants had suffered from minor injuries, six from moderate injuries, eight from serious injuries, and in eleven cases a family member pursued a claim following the death of the patient. Sixteen claimants successfully settled, and the remaining 14 withdrew following advice from their solicitor. Claimants were accessed via a law firm in a northern English city (Melville et al. 2014).
  The second study involved speaking to 30 claimants from across Scotland and was part of a project aimed at examining the potential impact of a no-fault scheme on claim outcomes and stakeholder satisfaction, including claimants (Stephen et al. 2012). Claimants were accessed via four law firms that have an accredited specialist in clinical negligence cases. Only five interviewees had received compensation, and no claims had progressed to a final hearing. Twenty interviewees were women, and there was a surprising geographical spread of claimants, considering that they were located via firms in Glasgow and Edinburgh.
  The interviewees covered a range of emotional issues. Most claimants had sued as they had suffered a very traumatic experience, such as the death of a child or a spouse. They also described feeling as if their concerns were dismissed by medical staff, being treated without respect and dignity, and struggling to discover what had gone wrong. The majority of claimants explained that they had sued in an effort to obtain an explanation and an apology, and to ensure that similar problems did not reoccur, rather than because they wanted financial compensation. Clinical negligence claimants do not have a high success rate, and many claimants were left feeling disappointed and disillusioned. Even claimants that obtained compensation felt that the outcomes that they wanted were not forthcoming, and that the money could not compensate for their sense of loss and grief. This sense of lack of closure was poignantly expressed by one claimant:
  

  …there was nothing involved for [the death of my son], which was the whole point of my claim. When I look at the money even for losing your son, I had to have a caesarean, so I’m now limited to the number of children that I can have, who’s to say I would ever have a boy again which I didn’t, I’ve had a baby girl… I’ll never take him to school, I’ll never see him walk, I’ll never share his first Christmas, I’ll never see him get married, things like that, nothing like that is taken into consideration. It’s all about what happened on the actual day but nothing kind of about what happens afterwards, so it’s all very, well there’s just no emotion when you go to court for things like this (female claimant, death, settled).

  
  The sensitive nature of the research impacted upon the conduct of the studies in a number of ways. First, privacy legislation meant that contact with claimants had to be made via the legal firms. Firms did not contact some clients whom they felt were too vulnerable to participate in the study, and this condition was also a requirement of ethics approval. These included some clients suffering from a terminal illness, those who had a psychiatric condition or who had gone through particularly traumatic experiences such as having a baby die. In addition, one firm did not contact claimants who were especially disgruntled with the legal services they received. The firms did not provide information concerning how many claimants were contacted and how many were removed from the sample, and thus the research had both an unknown response rate and an unknown response bias. This problem further highlights the impact of gatekeepers upon sensitive research.
  Recruitment of claimants involved the law firm sending out an information sheet. The information sheet explained the purpose of the research, the topics that were to be covered and the interview, although it did not explicitly inform potential participants of the possibility that the interviews may raise emotional responses. If claimants agreed to participate in the research, they signed a consent form, which was then returned via a replied paid envelope. This type of recruitment method produces notoriously low response rates as well as issues with response bias (Etter & Pemeger 1997). To boost response rates, participants would normally be sent pre-contact information (e.g. pre-contact postcards) and follow-up reminders (Edwards et al. 2002). In this instance, however, it was felt that some claimants may find it emotionally difficult to hear from a research team. It was assumed that claimants who wanted to participate would be motivated to respond without additional contacts. It was considered important to respect this decision, and so pre-contact information and reminders were not sent out. No doubt this decision impacted upon the response rate, but the principles of respect and of reducing the risk of emotional harm outweighed other concerns.
  Several claimants rang immediately after receiving the recruitment material. They wanted to know how the interviews would be conducted, and the interview topics. These questions were thoroughly covered in the information sheet. Thus, it appeared that claimants wanted to have an initial conversation, perhaps wanting to see if they would be treated respectfully and that their views would be heard, rather than wanting further information. This seemed especially important for people who had felt that neither their doctor, the lawyer who represented their case, nor medical experts who assessed their medical files had listened to their perspective.
  Initially, the English interviews used a semi-structured format that focused on the claimants’ experiences of legal services, which was the main aim of the studies. However, it quickly became obvious that efforts at focusing the claimant on topics that were of interest to the researcher but was less relevant to the claimant, produced answers that were short and inarticulate. For instance, one claimant had provided a rich, long and uninterrupted narrative of the incidences, including misdiagnosis of cancer, but also feeling as if he and his family had been treated without dignity or respect, which led him to sue:
  

  RESPONDENT: I was unhappy with [my doctors] because the way I looked at it is if my GP would have caught the problem when it started, it would have been caught five months earlier. My argument was if that would have been diagnosed five months earlier would I still have had to have the operation? Would I still have had to have chemotherapy? Could I have the operation and no chemotherapy? I was looking at all the ifs and buts and I was also looking at if I wouldn’t have gone to that night doctor, if my sister wouldn’t have forced me to go to that night doctor whether I would have been alive. Because that weekend I do believe that I was going to die because my body was packing up on me, it was giving up so I blame the doctor for them four or five occasions. My Mam knew, an old age pensioner, she demanded to go with me the last time I went to see him. And I said “Mam I’m nearly fifty, I’m not ten years old” and she said “I just want to see the doctor, see what he’s got to say while I’m there.” So anyway she come along, she bullied her way into coming along with me. And the doctor was talking to me and my mother answered one of the questions and the doctor looked at my mother and said “I’m asking your son not you, your son’s there he’s not a child he’s old enough to answer his own questions.” So I wasn’t happy about that because he was being disrespectful to my mother. So I’ve got those two issues with the doctor. The Hospital, I was there twice and both times they turned round and said that it was constipation and the third time the night doctor. Only because my sister insisted that I had an x-ray that I got to have that x-ray. So I went to a solicitors and told them everything I’ve just told you, they took my case on…

  
  At this point of the interview, the interviewer interrupted the respondent to ask the name of the solicitor. This change of topic severely disrupted the narrative flow of the interview, and the respondent’s responses became short and lacking in detail. The respondent also became concerned that they needed to provide the interviewer with the information that they required. Only after the interview returned to the respondent’s account of what had gone wrong, did the interview resume any narrative flow:
  

  INTERVIEWER: Can you tell me the solicitors that you went to?
  RESPONDENT: Smith and Co.¹x This is a pseudonym.
  INTERVIEWER: Who at Smith and Co, how did you find Smith and Co?
  RESPONDENT: It was just a fluke really, somebody told me about them, hearsay, somebody told me that they were the biggest solicitors in [the city].
  INTERVIEWER: When you contacted Smith and Co did you meet with a solicitor specifically or did you speak with them on the phone?
  RESPONDENT: No I had a meeting with one of them, and they were on something silly like a hundred and twenty pounds an hour.
  INTERVIEWER: Do you remember the name of the solicitor that you met?
  RESPONDENT: The name of the person that I was speaking to? If you just bear with me a minute I can even give you that other doctors, I’m walking round with my mobile now so I don’t know if you’re going to start breaking up.

  
  A review of the initial interview transcripts revealed that the interview format was problematic, and it was decided that unstructured interviews that avoided imposing research questions were more appropriate.
  Most of the interviews were conducted over the telephone. Face-to-face interviews may have been more appropriate as they allow for a greater sense of rapport (Holbrook et al. 2003, Irvine et al. 2013). However, the research was constrained by funding and time limits, and in particular, the geographical spread of the Scottish claimants would have made face-to-face interviews impractical. Nevertheless, there was one instance in which the claimant expressed a strong desire for a face-to-face interview, and it was felt that it was important to do the interview in person.
  The university’s IRB scrutinised the project to ensure that participants were fully informed of the emotional risk of participation and were aware of the topics that were to be raised. Interviewers were also provided with a list of support services that could be used to refer a claimant should the need arise. However, the IRB did not express concern for the emotional well-being of the members of the research team, which again highlights the indifference of university culture to the potential risks to their research staff.
  It was clear that the research took an emotional toll on the interviewers and other members of the research team. The emotional impact appeared to be strongest when researchers felt a personal connection between their own lives and the claimant’s experiences. For instance, the researcher who did the bulk of the interviews is married and has two young children. The interview that seemed to evoke the strongest emotions for her involved a female claimant with two children of a similar age and who had lost her husband to cancer. The emotional impact of the study was not just limited to the interviews. One study involved analysing NHS Scotland’s claims database, which contained information about the type of injury that had been suffered. One researcher was expecting the birth of his first grandchild, and found that knowing that some claims involved injuries to babies or mothers during childbirth to be especially difficult.
  The other main emotional strain came from feeling emotionally overloaded after either conducting or transcribing repeated interviews in a short span of time. Both projects involved the main interviewer speaking to many claimants during a relatively short period, and in some instances, several interviews were scheduled for the one day. There was only one transcriber employed for both projects, who was also asked to type multiple transcripts quickly. Both the interviewer and the transcriber reported feeling emotionally drained by the experience, and in retrospect, sufficient time needs to be allowed so that research staff can come to terms with the difficult material. This is especially important when there are a considerable number of interviews.
  The research team relied on informal mechanisms to provide emotional support, largely consisting of informal conversations after especially emotionally difficult interviews. However, these measures were not always adequate as they left the interviewer and transcriber needing to ask for support, and heavy work schedules meant that other members of the team were not always available. In retrospect, formal mechanisms for support that are built into the project’s timetable and budget would have been more adequate.

• 9 Conclusion

  It is clear that sensitive research can pose risks, especially the risk of re-victimisation for participants, and vicarious traumatisation for researchers, transcribers and coders. Thus, it is important that researchers have the skills to be able to minimise and manage these risks. However, sensitive research is often carried out by postgraduate students, solo researchers and others who often have limited support. In addition, legal researchers may not necessary have adequate training in conducting empirical research. Thus, having access to advice from researchers who have previous experience in conducting sensitive research is essential.
  It is possible to glean a number of key messages of advice from reflections written by experienced researchers. These include needing to be aware of how ethical principles, such as informed consent and maintaining confidentiality, can be put into practice. Researchers need to be careful to ensure that vulnerable participants understand what a study entails, and that participation is truly voluntary. Researchers also need to understand the limits on confidentiality, including situations that may require disclosure and problems arising from working within small communities.
  Researchers also need to be adequately prepared before conducting sensitive interviews, and should provide for the possibility of a participant becoming distressed. They need to be flexible, prepared to face ethical dilemmas ‘on the run’ and to have considered whether they will provide intervention for a participant in need. In addition, researchers should be prepared for dealing with interview material that covers issues that are painful, embarrassing or shameful. Preparation should also consider ways to open up conversational space, so that participants’ feelings and experiences can be sensitively explored rather than avoided. Researchers also need to be prepared for feeling as if they want to reciprocate, and to decide whether it is appropriate to give back to a participant, and if so, in what form.
  These key themes all highlight that conducting sensitive research requires reflective research practice. This involves anticipating, as best as possible, issues before they arise, having contingency plans and being flexible. The development of reflective research skills is also more likely to assist a researcher to deal with the demands on conducting sensitive interviews compared with relying on formal ethical guidelines. One of the problems associated with relying too heavily on ethical guidelines is that they often focus exclusively on protecting research participants from harm, whereas researchers can also experience emotional stress from listening repeatedly and deeply to people’s narratives of trauma, loss, grief, hurt and shame.
  Support systems for researchers conducting sensitive interviews are predominantly informal and ad hoc, and will therefore provide only limited assistance. Researchers are advised to ensure that formal support systems are in place, and to ensure that support is built into research budgets and timetables. Researchers also need to be aware that such systems may not necessarily be a high priority for granting bodies, grant holders and supervisors. It has been suggested that the lack of formal support reflects a lack of awareness of risk within university cultures, which further suggests that there is a need for researchers to lobby universities, funders, research teachers, research managers and others responsible for protecting researchers. In addition, there is also a need for more researchers to publish reflections on their personal experiences of conducting sensitive interviews. These reflections add to our shared knowledge about research practice, encourage researchers to seek advice, and remind solo and early career researchers that they are not alone.

References

• G.G. Abel, J.V. Becker, M.S. Mittelman, J. Cunningham-Rathner, J.L. Rouleau & W.D. Murphy (1987) ‘Self-Reported Sex Crimes of Non-Incarcerated Paraphiliacs’ 2(1) Journal of Interpersonal Violence 1987, p. 3-25

• M. Agar The Professional Stranger. San Diego: Academic Press 1996.

• A. Alty & K. Rodham ‘The Ouch! Factor: Problems in Conducting Sensitive Research’ 8(2) Qualitative Health Research 2008, p. 275-282.

• S. Bahn & P. Weatherill ‘Qualitative Social Research: A Risky Business When it Comes to Collecting “Sensitive Data’” 13(1) Qualitative Research 2013, p. 19-35.

• R. Banaker & M. Travers Theory and Method in Socio-Legal Research. London: Hart Publishing 2005.

• J. Barker & S. Weller ‘Geography of Methodological Issues in Research with Children’ 3(2) Qualitative Research 2003, p. 207-227.

• B. Beale, R. Cole, S. Hillege, R. McMaster & S. Nagy ‘Impact of In-Depth Interviewing on the Interviewer: Roller Coaster Ride’ 6(2) Nursing and Health Sciences 2004, p. 141-147.

• B. Bentley & M. O’Conner ‘Conducting Research Interviews with Bereaved Family Carers: When Do We Ask?’ 18(3) Journal of Palliative Medicine 2014, p. 241-245.

• D. Beyleveld & D. Townend ‘When is Personal Data Rendered Anonymous? Interpreting Recital 26 of Directive 95/46/EC’ 6(2) Medical Law International 2004, p. 73-86.

• B. Black, C. Silver, D.A. Hyman & W.M. Sage ‘Stability, Not Crisis: Medical Malpractice Claim Outcomes in Texas, 1988–2002’ 2 Journal of Empirical Legal Studies 2005, p. 207-259.

• R.J. Blendon, M. Catherine, M. DesRoches, M. Brodie, J.M. Benson, A.B. Rosen, E. Schneider, D.E. Altman, K. Zapert, M.J. Herrmann & A.E. Steffenson ‘Views of Practicing Physicians and the Public on Medical Errors’ 347 New England J. of Medicine 2002, p. 1933-1940.

• M. Bloor, B. Fincham & H. Sampson ‘Unprepared for the Worst: Risks of Harm for Qualitative Researchers’ 5(1) Methodological Innovations Online 2010, p. 45-55.

• T. Booth & W. Booth ‘The Use of Depth Interviewing with Vulnerable Subjects: Lessons from a Research Study of Parents with Learning Difficulties’ 39 Social Science and Medicine 1994, p. 415-424.

• R.R. Bovbjerg & A. Bartow Understanding Pennsylvania’s Medical Malpractice Crisis: Facts about Liability Insurance, the Legal System, and Health Care in Pennsylvania. Philadelphia: Pew Charitable Trusts 2003

• T.A. Brennan, L.L. Leape, N.M. Laird, L. Hebert, A.R. Localio, A.G. Lawthers, J.P. Newhouse, P.C. Weiler & H.H. Hiatt ‘Incidence of Adverse Events and Negligence in Hospitalized Patients: Results of the Harvard Medical Practice Study I’ 324(6) New England Journal of Medicine 1991, p. 370-376.

• T.A. Brennan, C.M. Sox & H.R. Burstin ‘Relation Between Negligent Adverse Events and the Outcomes of Medical-Malpractice Litigation’ 335(26) New England Journal of Medicine 1996, p. 1963-1967.

• E.J. Bromet & J.M. Havenaar ‘Basic Epidemiological Approaches to Disaster Research’ in F.H. Norris, S. Galea, M.J. Friedman & P.J. Watson (eds) Methods for Disaster Mental Health Research. New York: Guildford Press 2006.

• E.J. Brown ‘Double Whammy: Accessing, Recruiting and Retaining the Hidden of the Hidden’ 2(1) Journal of Ethnicity in Substance Abuse 2003, p. 43-51.

• S. Brzuzy, A. Ault & E.A. Segal ‘Conducting Qualitative Interviews with Women Survivors of Trauma’ 12(1) Affilia 1997, p. 76-83.

• G. Burr ‘Unfinished Business: Interviewing Family Members of Critically Ill Patients’ 3(3) Nursing Inquiry 1996, p. 172-177.

• R. Campbell Emotionally Involved: The Impact of Researching Rape. New York: Routledge 2002.

• R. Campbell, A.E. Adams, S.M. Wasco, C.E. Ahrens & T. Sefl ‘Training Interviewers for Research on Sexual Violence: A Qualitative Study of Rape Survivors’ Recommendations for Interview Practice’ 15 Violence Against Women 2009, p. 595.

• H.J. Carmack & J.M. DeGroot ‘Exploiting Loss?: Ethical Considerations, Boundaries, and Opportunities for the Study of Death and Grief Online’ 68(4) OMEGA--Journal of Death and Dying 2013, p. 315-335.

• S.C. Charles, J.R. Wilbert & K.J. Franke ‘Sued and Nonsued Physicians’ Self-Reported Reactions to Malpractice Litigation’ 1(42) American Journal of Psychiatry 1985, 437-440.

• E. Cohn & E. Larson ‘Improving Participant Comprehension in the Informed Consent Process’ 39(3) Journal of Nursing Scholarship 2007, p. 273-280.

• J. Corbin & J.M. Morse ‘The Unstructured Interactive Interview: Issues of Reciprocity and Risks when Dealing with Sensitive Topics’ 9 Qualitative Inquiry 2003, p. 335-354.

• M. Cowburn ‘Confidentiality and Public Protection: Ethical Dilemmas in Qualitative Research with Adult Male Sex Offenders’ 11(1) Journal of Sexual Aggression 2005, p. 49-63.

• K.V. Cowles ‘Issues in Qualitative Research on Sensitive Topics’ 10(2) Western Journal of Nursing Research 1988, p. 163-179.

• L.G. Criscione, J. Sugarman, L. Sanders, D.S. Pisetsky & E.W. St Clair ‘Informed Consent in a Clinical Trial of a Novel Treatment for Rheumatoid Arthritis’ 49 Arthritis and Rheumatology 2003, p. 361-367

• T. Damianakis & M.R. Woodford ‘Qualitative Research with Small Connected Communities: Generating New Knowledge while Upholding Research Ethics’ 22(5) Qualitative Health Research 2012, p. 708-718.

• P. Danzon ‘The Frequency and Severity of Medical Malpractice Claims’ 27(1) The Journal of law & economics 1984, p. 115-148.

• S. Darra ‘Emotional Work and the Ethics of Novice Insider Research’ 13 Journal of Research in Nursing 2008, p. 251-261.

• L. Dawson & N.E. Kass ‘Views of US Researchers about Informed Consent in International Collaborative Research’ 61 Social Science and Medicine 2005, p. 1211-1222.

• A.S. Demi & N.A. Warren ‘Issues in Conducting Research with Vulnerable Families’ 17(2) Western Journal of Nursing Research 1995, p. 188-202.

• V. Dickson-Swift, E.L. James, S. Kippen & P. Liamputtong ‘Blurring Boundaries in Qualitative Health Research on Sensitive Topics’ 16(6) Qualitative Health Research 2006, p. 853-871.

• V. Dickson-Swift, E.L. James, S. Kippen & P. Liamputtong ‘Doing Sensitive Research: What Challenges do Qualitative Researchers Face?’ 7 Qualitative Research 2007, p. 327-353.

• V. Dickson-Swift, E.L. James, S. Kippen & P. Liamputtong ‘Researching Sensitive Topics: Qualitative Research as Emotional Work’ 9(1) Qualitative Research 2009, p. 61-79.

• A. Durham ‘Developing a Sensitive Practitioner Research Methodology for Studying the Impact of Child Sexual Abuse’ 32(4) British Journal of Social Work 2002, p. 429-442.

• P. Edwards, I. Roberts, M. Clarke, C. DiGuiseppi, S. Pratrap, R. Wentz & I. Kwan ‘Increasing Response Rates to Postal Questionnaires: Systematic Review’ 324(7347) British Medical Journal 2002, p. 1183-1186.

• M. Ennis & C. Vincent ‘The Effects of Medical Accidents and Litigation on Doctors and Patients’ 16(2) Law & Policy 1994, 97-121.

• J. Etter & V.T. Perneger ‘Analysis of Non-Response Bias in a Mailed Health Survey’ 50(10) Journal of Clinical Epidemiology 1997, p. 1123-1128.

• R.R. Faden, T.L. Beauchamp & N.M. King A History and Theory of Informed Consent. Oxford: Oxford University Press 1986.

• K. Featherstone & J.L. Donovan ‘Random Allocation or Allocation at Random? Patients’ Perspectives of Participation in a Randomised Controlled Trial’ 317 British Medical Journal 1998, p. 1177-1180.

• N. Findholt & L.C. Robrecht ‘Legal and Ethical Considerations in Research with Sexually Active Adolescents: The Requirement to Report Statutory Rape’ 34(5) Perspectives on Sexual and Reproductive Health 2002, p. 259-264.

• H. Genn, M. Partington & S. Wheeler Law in the Real World: Improving our Understanding of How Law Works. London: Nuffield Foundation 2006.

• K.R. Gilbert ‘Collateral Damage? Indirect Exposure of Staff Members to the Emotions of Qualitative Research’ in: K.R. Gilbert (ed.) The Emotional Nature of Qualitative Research. Boca Raton: CRC Press 2001a, p. 147-161.

• K.R. Gilbert ‘Introduction: Why are We Interested in Emotions?’ in: K.R. Gilbert (ed.) The Emotional Nature of Qualitative Research. Boca Raton: CRC Press 2001b, p. 3-15.

• M. Glazer ‘The Threat of the Stranger’ Hastings Center Report 1980, p. 25-31.

• S. Goodrum & J.L. Keys ‘Reflections on Two Studies of Emotional Sensitive Topics: Bereavement from Murder and Abortion’ 10(4) International Journal of Social Research Methodology 2007, p. 249-258.

• S. Halliday & P. Schmidt Conducting Law and Society Research: Reflections on Methods and Practices. Cambridge University Press: Cambridge 2009.

• H.H. Hiatt, B.A. Barnes, T.A. Brennan, N.M. Laird, A.G. Lawthers, L.L. Leape, A.R. Localio, J.P. Newhouse, L.M. Peterson, K.E. Thorpe & P.C. Weiler ‘A Study of Medical Injury and Medical Malpractice’ 321 New England J. of Medicine 1989, p. 480-484.

• G.B. Hickson, E.W. Clayton, P.B. Githens & F.A. Sloan ‘Factors that Prompted Families to File Medical Malpractice Claims Following Perinatal Injuries’ 267(10) JAMA 1992, p. 1359–1363.

• J.M. Hirshon, S.D. Krugman, M.D. Witting, J.P. Furuno, M.R. Limcangco, A.R. Perise & E.L. Rasch ‘Variability in Institutional Review Board Assessment of Minimal Risk Research’ 9 Academic Emergency Medicine 2002, p. 1417-1420.

• A.L. Holbrook, M.C. Green & J.A. Krosnick ‘Telephone Versus Face-to-Face Interviewing of National Probability Samples with Long Questionnaires: Comparisons of Respondent Satisficing and Social Desirability Response Bias’ 67(1) Public Opinion Quarterly 2003, p. 79-125.

• J.S. Horner ‘Retreat from Nuremberg: Can we Prevent Unethical Medical Research’ 113(5) Public Health 1999, p. 205-210.

• K. Hoeyer, L. Dahlager & N. Lynöe ‘Conflicting Notions of Research Ethics: The Mutually Challenging Traditions of Social Scientists and Medical Researchers’ 61(8) Social Science & Medicine 2005, p. 1741-1749.

• G. Hubbard, K. Blackett-Milburn & D. Kemmer ‘Working with Emotional Issues for the Researcher in Fieldwork and Teamwork’ 4(2) International Journal of Sensitive Research Methodology 2001, p.119-137.

• L.R. Huycke & M.M. Huycke ‘Characteristics of Potential Plaintiffs in Malpractice Litigation’ 120 Annals of Internal Medicine 1994, pp. 792–798.

• A. Irvine, P. Drew & R. Sainsbury ‘“Am I Not Answering Your Questions Properly?” Clarification, Adequacy and Responsiveness in Semi-Structured Telephone and Face-to-Face Interviews’ 13(1) Qualitative Research 2013, p. 87-106.

• M. Israel ‘Strictly Confidential? Integrity and the Disclosure of Criminological and Socio-Legal Research’ 44 British Journal of Criminology 2004, p. 715-740.

• M. Israel & I. Hay Research Ethics for Social Scientists: Between Ethical Conduct and Regulatory Compliance. London: Sage Publications 2006.

• M.A. Jacob ‘Form-Made Persons: Consent Forms as Consent’s Blind Spot’ 30(2) PoLAR 2007, p. 249-268.

• R. Jagsi, B.T. Kitch, D.F. Weinstein, E.G. Campbell, M. Hutter & J.S. Weissman ‘Residents Report on Adverse Events and their Causes’ 165(22) Archives of Internal Medicine 2005, p. 2607-1613.

• M. Jefford & R. Moore ‘Improvement of Informed Consent and the Quality of Consent Documents’ 9(5) The Lancet Oncology 2008, p. 485-493.

• B. Johnson & J. Macleod Clarke ‘Collecting Sensitive Data: The Impact on Researchers’ 13 Qualitative Health Research 2003, p. 421-434.

• S. Kvale Interviewing: An Introduction to Qualitative Research Interviewing. London: Sage Publications 1996.

• E. Larson, T. Bratts, J. Zwanziger & P. Stone ‘A Survey of IRB Process in 68 U.S. Hospitals’ 36 Journal of Nursing Scholarship 2004, p. 260-264.

• R.M. Lee Doing Research on Sensitive Topics. London: Sage 1993.

• R.M. Lee & C.M. Renzetti ‘The Problems of Researching Sensitive Topics’ 33 American Behavioural Scientist 1990, p. 510-528.

• P. Liamputtong Researching the Vulnerable: A Guide to Sensitive Research Methods. London: Sage Publications 2007.

• C. Malacrida ‘Reflexive Journaling on Emotional Research Topics: Ethical Issues for Team Researchers’ 17(10) Qualitative Health Research 2007, p.1329-1139.

• N.A. McWilliams, E.L. Jones & R.J. Davies ‘Problematic Variation in Local Institutional Review of a Multicenter Genetic Epidemiology Study’ 290 JAMA 2003, p. 360-366.

• A.L. Melville, F.H. Stephen & T. Krause ‘“He Did Everything He Possibly Could For Me”: Medical Malpractice Claimants’ Experiences of Lawyer–Client Relations’ 21(2) International Journal of the Legal Profession 2014, 171-193.

• D.M. Mertens & P.E. Ginsberg ‘Deep in Ethical Waters: Transformative Perspectives for Qualitative Social Work Research’ 7 Qualitative Social Work 2008, p. 484-503.

• M. Moore, M. Maple, AM Mitchell & J Cerel ‘Challenge and Opportunities for Suicide Bereavement Research: The Experience of Ethical Review’ 34(5) Crisis: the Journal of Crisis Intervention and Suicide Prevention 2013, 297-304.

• V. Morrow & M. Richards ‘The Ethics of Social Research with Children: An Overview’ 10(2) Children and Society 1996, p. 90-105.

• E. Murphy & R. Dingwall ‘The Ethics of Ethnography’ in: P. Atkinson, A. Coffey, S. Delamont, J. Lofland & L. Lofland (eds), Handbook of Ethnography. London: Sage Publications 2001.

• L. Nash, M. Daly, M. Johnson, G. Walter, M. Walton, S. Willcock, C. Coulston, E. van Ekert, & C. Tennant ‘Psychological Morbidity in Australian Doctors Who Have and Have Not Experienced a Medico-Legal Matter: Cross-Sectional Survey’ 41(11) Australian and New Zealand Journal of Psychiatry 2007, 917-925.

• E. Owens ‘Conversational Space and Participant Shame in Interviewing’, 12(6) Qualitative Inquiry 2006, p. 1160-1179.

• M.K. Paasche-Orlow, H.A. Taylor & F.L. Brancati ‘Readability Standards for Informed-Consent Forms as Compared with Actual Readability’ 348(8) New England Journal of Medicine 2003, 721-726.

• T. Relis ‘It’s Not About the Money! A Theory on Misconceptions of Plaintiffs’ Litigation Aims’ 68(2) University of Pittsburgh Law Review 2006, p. 341-385.

• M. Rice & M.E. Broome ‘Incentives for Children in Research’ 26(2) Journal of Nursing Scholarship 2004, p. 167-172.

• P.C. Rosenblatt ‘Ethics of Qualitative Interviewing with Grieving Families’ 19 Death Studies 1995, p. 139-155.

• L. Sanford Strong at the Broken Places. London: Virago 1990.

• B. Saunders, J. Kitzinger & C. Kitzinger ‘Anonymising Interview Data: Challenges and Compromise in Practice’ 15(5) Qualitative Research 2015, p.616-632.

• R. Scarce ‘(No) Trial (But) Tribulations: When Courts and Ethnography Conflict’ 23 Journal of Contemporary Ethnography 1994, p. 123-49.

• D. Scully Understanding Sexual Violence: A Study of Convicted Rapists. Cambridge: Mass.: Unwin Hyman 1990.

• R.S. Shapiro, D.E. Simpson, S.L. Lawrence, A.M. Talsky, K.A. Sobocinski & D.L. Schiedermayer ‘A Survey of Sued and Nonsued Physicians and Suing Patients’ 149(10) Archives of Internal Medicine 1989, p. 2190–2196.

• E. Sherry ‘The Vulnerable Researcher: Facing the Challenges of Sensitive Research’ 13(3) Qualitative Research Journal 2013, p. 278-288.

• F.A. Sloan & C.R. Hsieh (1995) ‘Injury, Liability, and the Decision to File a Medical Malpractice Claim’ 29(3) Law & Society Review 1995, p. 413–435.

• F.H. Stephen, A.L. Melville, & T. Krause. A Study of Medical Negligence Claiming in Scotland. Edinburgh: Scottish Government Social Research 2012.

• M.M. Stevens, B.A. Lord, M-T. Proctor, S. Nagy & E. O’Riordan ‘Research with Vulnerable Families Caring for Children with Life-Limiting Conditions’ 20 Qualitative Health Research 2009, p. 496-505.

• D.M. Studdert, M.M. Mello, A.A. Gawande, T.K. Gandhi, A. Kachalia, C. Yoon, A.L. Puopolo & T.A. Brennan ‘Claims, Errors, and Compensation Payments in Medical Malpractice Litigation’ 354(19) New England Journal of Medicine 2006, p. 2024-2033.

• J.Y. Taylor ‘Talking Back: Research as an Act of Resistance and Healing for African American Women Survivors of Intimate Male Partner Violence’ 25(3) Women & Therapy 2002, p. 145-160.

• C.R. Trulson, J.W. Marquart & J.L. Mullings ‘Breaking in: Gaining Entry to Prisons and Other Hard-to-Access Criminal Justice Organizations’ 15(2) Journal of Criminal Justice Education 2004, p. 451-478.

• J.E. Turnball, J.D. McLeod, J.M. Callahan & R.C. Kessler ‘Who Should Ask? Ethical Interviewing in Psychiatric Epidemiology Studies’ 58(2) American Journal of Orthopsychiatry 1988, p. 228-239.

• N. Vidmar, P. Paul, K. MacKillop & K. McCarthy ‘Uncovering the Invisible Profile of Medical Malpractice Litigation: Insights from Florida’ 54 DePaul Law Review 2004, p. 315-356.

• C. Vincent, A. Phillips & M. Young ‘Why Do People Sue Doctors? A Study of Patients and Relatives Taking Legal Action’ 343(8913) The Lancet 1994, p. 1609–1613.

• J.H. Watts ‘Emotion, Empathy and Exit: Reflections on Doing Ethnographic Qualitative Research on Sensitive Topics’ 3(2) Medical Sociology Online 2008, 3-14.

• R. Wiles, G. Crow, S. Heath & V. Charles ‘The Management of Confidentiality and Anonymity in Social Research’ 11(5) International Journal of Social Research Methodology 2008, p. 417-428.

• D.A. Wirshing, W.C. Wirshing, S.R. Marder, R.P. Liberman & J. Mintz, J. ‘Informed Consent: Assessment of Comprehension’ 155(11) American Journal of Psychiatry 1998, p.1508-1511.

• L.L. Woodby, B.R.Williams, A.R. Wittich & K.L. Burgio ‘Expanding the Notion of Researcher Distress: The Cumulative Effects of Coding’ 21(6) Qualitative Health Research 2011, p. 830-838.

• I.M. Wong ‘The Ethics of Rapport: Institutional Safeguards, Resistance, and Betrayal’ 4 Qualitative Inquiry 1998, p. 178-199.

Noten

• 1 This is a pseudonym.