“What Should I Do Now?” 1 : Navigating Relational Ethics in Practice as an Early Career Researcher as Illustrated by a Qualitative Interview Study About Women’s Drinking Practices

Abstract

Qualitative interview studies on sensitive topics often draw on principles of feminist methodologies which focus on developing and maintaining non-exploitative, caring relationships with participants. For early career researchers, who may have less research experience, managing relational ethical issues that arise in research relationships can be difficult. Additionally, they could experience further pressures because of their junior roles and precarious employment. In the context of health research, early career researchers working on qualitative studies may experience specific challenges because of the predominance of the biomedical paradigm in this discipline. In this article, I explore some of the relational ethical issues I deliberated as an early career researcher when working in a medical faculty on a semi-structured qualitative interview study about women’s alcohol drinking practices. I focus on two overlapping themes from my experience of ethics in practice “Trying to building and maintain relationships” and “Trying to stabilize inequalities in research relationships.” With a primary focus on how I negotiated differing responsibilities, I draw on examples from the community-based face-to-face and virtual recruitment, the fast-paced face-to-face interviews, and the process of returning interview transcripts to women to review. With this analysis, I contribute to existing literature about ethics in practice for early career researchers by indicating the types of relational ethics that will need to be navigated and the resources needed to support them. These resources include having adequate time, opportunities for reflection, and good supervisory support. I also contribute to scholarship which critiques the wider health research context by considering the challenges it can pose for early career researchers when managing relational ethical spaces in qualitative interview studies. This article will be of relevance to novice researchers and those who supervise and manage them.

Introduction

Ethical considerations are part of all research, but one-off, semi-structured interviews, the most commonly used qualitative method (Allmark et al., 2009), are sometimes viewed as less troublesome because they are more bounded than ethnographic or participatory methods (Miller, 2017). However, qualitative interview studies which investigate sensitive aspects of people’s lives often draw on principles from the broad range of feminist methodologies (Kvale, 2007; Miller, 2017; Sampson et al., 2008), which include attentiveness to ethical issues in research and a sensitivity to the unequal relationships between researchers and participants in the research process (Hesse-Biber, 2014; Letherby, 2003; O’Shaughnessy & Krogman, 2012; Ramazanoglu & Holland, 2002). Therefore, due to the close social interactions between researchers and participants, qualitative interview studies can raise complex relational ethical issues to respond to (Allmark et al., 2009; Miller, 2017). Early career researchers frequently carry out the fieldwork for one-off interview studies on very personal and sensitive topics (Mauthner & Doucet, 2008) and may be unprepared for the hidden relational ethical issues that they present. Detailed accounts of ethics in practice from researchers, who use interviews, are needed to highlight and signal to early career researchers, and those who supervise and manage them, the ethical deliberations managed in the spaces outside of the procedural ethical review (Edwards & Mauthner, 2012; Wiles & Boddy, 2013). Thus, in this article, I will candidly describe the ethical issues I navigated in practice, as an early career researcher working within a contemporary academic health research context, on a semi-structured interview study about women’s alcohol drinking practices and stress. Using this example, I contribute to scholarship about relational ethics in practice. I also extend debates about the obstacles posed by the health research context for qualitative researchers, by illustrating some challenges it can present for supporting early career researchers when negotiating relational ethics in interview studies that require close interactions.

Early Career Researchers and the Contemporary Health Research Context

The term early career researcher (hereon referred to as ECR) is currently defined in different ways and should ideally recognize individuals’ circumstances (Bosanquet et al., 2017). For consistency, I will use the term ECR to include PhD researchers and postdoctoral researchers in the first 5 years after they have completed their doctorate; PhD researchers are training in research methods, and postdoctoral researchers are often the most junior members of research teams (Mauthner & Doucet, 2008; Rimando et al., 2015). The contemporary academic context in the UK, as in other European and Western countries, is characterized by competition which rewards rapid productivity (Caretta et al., 2018; do Mar Pereira, 2016; Gill, 2009; Webster et al., 2019). This working environment places pressures on researchers and academics at all career levels but can be especially challenging for ECRs (Bosanquet et al., 2017; Caretta et al., 2018) who often have precarious contracts and face pressure to finish fieldwork and publish their findings by the end of a fixed term (Batty, 2018; Caretta et al., 2018; Nature Jobs, 2017). These pressures experienced by ECRs, together with their limited experience of fieldwork, might make responding with integrity during fieldwork more difficult (UK Research and Innovation, 2020).

Different academic disciplines also present varying challenges to working with integrity (UK Research and Innovation, 2020). My focus in this article is the health research context. Although qualitative research is now commonly used in health research, the discipline remains dominated by a biomedical and post-positivist research paradigm where quantitative research is the norm (Eakin, 2016; Kontos & Grigorovich, 2018; Webster et al., 2019). Other authors have described how qualitative researchers, working in health settings, can find themselves needing to fit their work into frameworks which may not be suitable for the contemplative and often slow inquiry which is needed for qualitative research (Burns et al., 2018; Eakin, 2016; Kontos & Grigorovich, 2018; Webster et al., 2019). Qualitative researchers may also find that the work they do is misunderstood, or that the interpersonal skills needed for managing the relational aspects of ethical qualitative research are underestimated (Eakin, 2016). During the study I describe here I was a mature, part-time PhD student in a health services research institute in a medical faculty and was supervised by a multi-disciplinary team of three senior academics. Two of my supervisors were based in the institute where I was based, and the other was in the faculty of social sciences and oversaw the qualitative fieldwork.

Ethical Principles and Relational Ethics in Practice

Research ethics concerns the moral process of considering the wellbeing of others when conducting studies (Boulton, 2009). Common principles of research ethics include requesting that participants give informed consent, ensuring no harm comes to participants through the research, and guaranteeing their confidentiality and anonymity (Boulton, 2009; Pollock, 2012). In the UK, health research involving humans requires a procedural ethical review before the study begins. Similar structures operate in Australia, America, Canada, and Europe (Pollock, 2012). This framework requires researchers to consider how these fundamental ethical principles will be applied to the methods, populations, and topics of their study. The initiation of a study is then dependent on the approval of these decisions by an Institutional Review Board or a Research Ethics Committee (Pollock, 2012). My primary focus here is the relational spaces that ECRs need to navigate, alone and with integrity, when conducting fieldwork which is best considered through ethics in practice.

Ethics in practice concerns researchers’ situated responses to issues during the research process (Guillemin & Gillam, 2004). This approach recognizes that researchers, like participants, are embedded in relationships with their own values and responsibilities which can alter how they deliberate and respond to different ethical situations (Banks et al., 2013). This approach also asserts that there is usually more than one solution to an ethical problem (Holland et al., 2014). As such, when conducting fieldwork, researchers’ tend to consider the question of, “what should I do now?” rather than, “this is what you should do now,” as the procedural approach alone might be seen to advocate (Ellis, 2007, p. 4). Overall, considering ethics in practice underlines that, while it is always a priority for researchers, making decisions about the wellbeing of participants during qualitative fieldwork is dependent on various situated factors (Edwards & Mauthner, 2012; Hallowell et al., 2005).

A key element of the relational ethical dynamic is concerned with how researchers negotiate various responsibilities in the research process (Sampson et al., 2008). For example, they need to strike a balance between caring for the participants, answering the research questions, and other responsibilities they may have such as their own family commitments. Undoubtedly, conflicts between different duties and obligations can be extremely difficult for researchers to navigate and a key skill for researchers is ensuring the wellbeing of the participants while managing a range of other concerns (Boulton, 2009; Hallowell et al., 2005). From here I will introduce the topic, methodology, and method of the study, and then go on to consider ethics in practice.

Research Topic, Methodology and Method

The study set out to explore women’s non-dependent alcohol consumption in relation to stress and aimed to contribute critical sociological insights to understanding this phenomenon. Although non-dependent drinking practices are relatively normalized among many groups of women in the UK, other qualitative research has illustrated that it is still a gendered and moral practice (Emslie et al., 2015; Nicholls, 2019). I therefore, wanted to adopt a methodology and method of data collection that was suitable and appropriate for studying a private aspect of a woman’s life which may not be commonly discussed (Dickson-Swift et al., 2008).

I adopted key principles of feminist methodologies because as noted above, they are closely associated with studying sensitive topics, and other researchers’ have called attention to the value of doing so when studying and theorizing women’s drinking practices (e.g. Ettorre, 2018; Staddon, 2015). The method of qualitative, semi-structured interviews was chosen for the data collection as it was hypothesized that it would help to illuminate aspects of women’s alcohol consumption and stress that were most important to them (Doucet & Mauthner, 2008; Kvale, 2007).

Before beginning the fieldwork, I was granted ethical approval by the Newcastle University Faculty of Medical Sciences Ethics Committee (00443/2011). This procedural review had encouraged me to consider how ethical principles should be applied to the project. Women were eligible to take part in the study if they were: (i) of legal drinking and working age in the UK (18–64 years), (ii) were not in treatment for an alcohol use disorder, (iii) had ever drunk alcohol, and (iv) lived in the North East of England. It was assumed that all women would have experienced “stress,” which I approached as a range of discourses, drawn on to make sense of environmental demands and adverse experiences (Harkness et al., 2005). I also wanted to ensure I was incorporating the views of women from working-class and middle-class communities which I defined using an area deprivation measure.

The fieldwork took place over 1 year, between May 2014 and June 2015. In this year, I recruited women using several strategies of community-based recruitment. In total, I undertook face-to-face, one-off, semi-structured interviews with 28 women in their homes (n = 14), community venues (n = 11), or workplaces (n = 3). Twenty-six of these interviews were included in the final sample. One of the key findings of the study have been reported elsewhere (Jackson et al., 2018).

In the next section, I focus on two cross-cutting and related themes from my experiences of ethics in practice “Trying to build and maintain relationships” and “Trying to stabilize inequalities in research relationships.” Across both themes, I aimed to emphasize the negotiations involved in managing different responsibilities in the research process. Some examples also illustrate emotional labor which is another theme widely discussed by other researchers (e.g. Batty, 2018; Carroll, 2013; Dickson-Swift et al., 2009; Hallowell et al., 2005). The examples include everyday ethical deliberations and two “ethical speedbumps” which Holland et al. (2014) refer to as moments that bring ethical issues to the fore and cause the researchers to respond sensitively and quickly. I use anonymized participant data and the project fieldwork diary to illustrate some of the examples. All the participants’ names have been changed.

Ethics in Practice

Trying to Build and Maintain Relationships

Feminist methodologists have helped to show that building good relationships and developing rapport with participants in an understanding, empathetic, and engaging manner is more suitable for researching sensitive topics than asymmetrical approaches associated with the post-positivist tradition (Duncombe & Jessop, 2012; Prior, 2018; Sampson et al., 2008). Work from feminist scholars has also illustrated that developing close relationships and creating conditions for participants to speak about sensitive and personal aspects of their lives can raise a range of ethical concerns in the process of the research, which often cannot be anticipated or planned for (Doucet & Mauthner, 2008; Duncombe & Jessop, 2012; Kirsch, 2005). In this section, I focus on everyday relational ethical issues I navigated while developing and trying to maintain relationships during the recruitment and the interviews.

Building rapport during face-to-face and virtual recruitment

The recruitment stage of an interview study is vital for building relationships with participants before the interviews take place. However, the work that is involved in developing these connections is often unseen (Miller, 2017). The community and virtual recruitment methods used in this study raised different challenges to respond to. I began recruiting face-to-face in community groups as I felt that seeing and talking to me would help to build trust prior to the interviews. In some community groups the group leaders, who acted as gatekeepers, played a crucial role in facilitating my relationships with the groups of women that they worked with (McAreavey & Das, 2013). In one setting, a school social group, the group leader mentioned that some of the women lacked trust in professionals due to difficulties that they had experienced in their lives. In addition, I was made aware that others had previously experienced the negative effects of very heavy drinking in their childhood. Before granting me access, the group leader reassured the group that I was not there to check up, monitor, or judge their alcohol consumption.

On my first visit to this school social group, I explained the study informally to the women. I then found, as I did in other groups, that speaking to the women about common interests and joining in with their activities, helped to build good relationships. The formal written study participant information sheet was always given out, but was often side-lined as I felt it drew attention to my professional role and did not feel compatible with making the women feel comfortable (Waycott et al., 2015). During my first visit to this group, some women said they would like to take part in an interview and that I could come back over the following weeks to interview them in the group leader’s private office. In the procedural ethical review for the study, I had said I would try to interview participants’ in their homes as the home environment is often perceived to make participants feel comfortable, and can help researchers’ gain a better insight into the participants’ private lives (Britten, 1995). However, in practice, the women suggested they would be more at ease in the familiar surroundings of the school as they were protected by their peers and their group leader’s presence.

Overall, I visited this group five times over 2 months and, although it can sometimes be assumed that familiarity with the participants can make them feel more comfortable, in practice the women’s increased familiarity with me seemed to work in different ways in terms of them wanting to take part in the research. One woman, who during my first visit had referred to me as the “drug awareness woman,” had said she did not want to take part. However, on a later visit stated that she would like to take part. This may be because, over time, she got to know me better, and had seen some of her friends taking part. Conversely, another woman who, on my first visit was particularly welcoming and had said she wanted to participate in the study, said on a later visit that she no longer wanted to. She did not tell me explicitly why she chose not to participate, but it may be that she did not want to talk to me about drinking and stress in her life after we had got to know each other quite well. Indeed, feminist scholarship has illustrated that familiarity between participants and researchers is not straightforward and that sometimes it can inhibit involvement in a research study; a researcher’s position as a stranger can sometimes make it easier for people to talk about sensitive issues (Couture et al., 2012; Letherby, 2003). Overall, these examples point to the types of everyday ethical and practical dynamics that emerged in the five, different community-based recuitment settings. I needed to be flexible and respond to the women’s individual circumstances to feel that I was caring for them in the best way I could.

Throughout the fieldwork, I was also working on another component of the study, a qualitative systematic review, which had delayed the start of the recruitment as it had taken longer than expected at the beginning of the project. Though I had initially planned for all the recruitment to be face-to-face in community groups, this method of recruitment was quite labor intensive. Accordingly, once the interviews started, I modified my plan and began using the social media site, Facebook, to support the community group recruitment. This virtual recruitment was initially perceived as a less labor-intensive method of recruitment, but in actuality, I found it raised different practical and ethical issues to respond to (Fileborn, 2016).

The Facebook page for the study, gave potential participants limited information about the project, the option of viewing the dedicated website and an email address if they wanted further information about taking part. When I replied to potential participants who contacted me virtually, I sent a copy of the study participant information sheet and asked them to tell me a bit about themselves so that I could determine if they were eligible for the study. As mentioned above, the study excluded women who were in treatment for heavy alcohol use. However, because I wanted to maintain the informal tone of the email, I did not explicitly ask them if they had been in treatment. Instead, at this stage, I relied on them to read the participant information sheet to determine for themselves whether they were eligible. This resulted in one of the more memorable “ethical speedbumps” in the study with Lorna, a participant I had recruited through the Facebook page.

I had emailed Lorna before her interview and arranged to meet her at the home of the relative she was living with. When I arrived, Lorna seemed pleased to see me. As I had recruited her through social media, I knew less about her than I did some of the participants who I had recruited face-to-face. When we began speaking, she told me quite quickly that she was in treatment for alcohol use. As I had not explicitly mentioned this aspect of the eligibility criteria in the email, I had to quickly decide to either tell her she was not eligible because she was in treatment, or to carry on with the interview. The decision I made was to carry on and reduce the number of questions posed to her. I felt that telling Lorna she was ineligible in front of her family and in her home, could have upset her. My fieldwork notes, made a few hours after the interview, illustrate this:

Lorna wanted to help so I continued the interview. I tried to ask a few of the questions, but it was obviously painful for her so I didn’t carry on asking about this. Maybe I should have stopped sooner or said that I didn’t think she was in the right place. But because she wanted to help, I continued. I am worried, but she has her [relative] to support her who is aware of her circumstances. (Fieldnotes, November 2014)

After the interview, Lorna and I went to a local café and I felt that the experience of talking to me, particularly after the interview, might have been temporarily beneficial for Lorna as she seemed lonely. On reflection, I personally felt that my decision to continue was the best balance between caring for Lorna and caring for myself; I was trying to protect Lorna from becoming upset by not telling her that she was ineligible and I was caring for myself by avoiding the anxiety that may have arisen because the eligibility criteria was not made clearer at the start. This may not be the decision I would make again but, at the time, it felt like the right thing to do.

My qualitative supervisor had told me I could contact her if any difficult issues came up during the fieldwork. Thus, after the interview with Lorna I contacted my supervisor at the earliest opportunity. I asked for reassurance that I had done what I could to protect and care for Lorna. This was an embarrassing situation, but I felt comfortable talking to my supervisor and I knew that she would understand. However, as Lorna had access to all the relevant participant materials to make an informed decision about taking part and had highlighted that she understood them during the consent process, guidance stated that these issues should be resolved internally and therefore, it was not reported back to the research ethics committee. Regardless of this outcome, I agreed with my supervisor that in the future I should take care to explicitly ask whether the participants were receiving formal treatment for heavy alcohol use, rather than relying on them reading the participant information. We also agreed that I should introduce a phone call into the recruitment process to find out a bit more about women’s experiences before the interviews. My supervisor’s guidance and expertise enabled me to see that I needed to gain a better balance between the informal and professional relationship and take more control over the recruitment to support the women and myself. This was also made evident by Kirsch (2005) who noted that one of the challenges of using principles of feminist methodology, for novice researchers, is that when trying to build relationships, other important ethical considerations can sometimes be overlooked. I did not inform Lorna that her interview was not included in the final analysis because I felt it may upset her to do so. This interview, along with other informal sources, has been important in drawing my attention to wider issues relating to women’s heavy drinking, which is something I am careful to reflect on in other outputs from the project.

Managing responsibilities during interviews

The focus on stress in this study meant I had expected to discuss emotional topics in the interviews. In hindsight however, at the start of the project, I was unprepared for range of sensitive and emotional experiences that would be discussed. A particular ethical challenge in practice was working out how to balance caring for the women in the fast-paced interactions, with keeping focused on the research questions.

An example of how I thought about and navigated these different concerns is demonstrated by the interview with Jane who I recruited from a community group. In the extract below she was discussing her partner’s drinking and touched on the fact that there had been violence related to very heavy drinking in her childhood:

Jane:

I’m not an aggressive drunk. Like

Katherine:

No

Jane:

Na, na. Being around that when I was growing up, and I’m not like that at all. I’m not aggressive. Not at all. Not at all, and that’s why I think I’m glad that Mark doesn’t drink either…Because we’re both not drunk, so it’s all just about, crack on, and he brings us home and puts us to bed, bless him. (Laughter)

Katherine:

(Laughter) He sounds great!

I did not follow up on what Jane said about her father’s drinking here or later on in the conversation. Instead, I made a joke to refocus the conversation on what Jane said about her partner, Mark. Not addressing this part of Jane’s life was intentionally done to protect her from any harm being brought about by discussing experiences not directly related to the focus of the interviews. I was also caring for myself, as it may have increased the emotional labor for me if Jane had become upset. Moreover, if she had begun to talk about her experiences, it may have distracted from the wider focus of the interview, which I only had limited time to manage. On reflection, while I had no sense that Jane and other participants did not feel cared for, they may have seen my lack of engagement with these aspects of their life as not acknowledging their experiences as important. Jane may have understood why I did not explore it further, but this was left unsaid. Many other researchers have noted that caring for participants while focusing on the research question can be one of the most difficult relational aspects to manage in interview studies. Often, it is also made more difficult because of the limited time there is to engage with participants (Cooper & Rogers, 2015; Miller, 2017; Roberts, 2019).

There were many times during the fieldwork when I did follow up on what women said about sensitive topics, or without prompting them to do so, they discussed them with me. I found, as other researchers have described, that the rapport or the opportunity to talk about their lives could encourage women to disclose deep and private things that they may not have expected to share (e.g. Dickson-Swift et al., 2009; Duncombe & Jessop, 2012; Kirsch, 2005). An example of this was with Ruby, in one of the few interviews that took place in a workplace. She spoke about a mental health condition that she was dealing with and had previously received treatment for. After the interview, it was clear the conversation had involved emotional labor for both of us and we were able to spend some time together discussing other things. She mentioned that she had not spoken to many people about this part of her life and said that she had not expected to discuss it in the interview. I emailed her later to see how she was after the interview. In her reply, she also asked me if I was alright and mentioned it must have been a difficult interview to manage. This illustrates the complexity of the relational dynamic in interviews and how participants sometimes care for researchers too (Hallowell et al., 2005).

Feminist and other qualitative methodologists continue to debate whether researchers should engage with the very difficult aspects of people’s lives during interviews or whether they should try to avoid very sensitive topics if they are not directly related to the focus of research projects (Kirsch, 2005; O’Shaughnessy & Krogman, 2012; Shaw, 2011). Some scholars argue that participants should be made aware that the interview is not a therapeutic encounter and that the researcher may not have the skills to support them (Kirsch, 2005). Writing about narrative approaches, Hollway and Jefferson (2013) make the point that it could be reassuring for participants to speak about a traumatic event in a safe context. Other researchers suggest that avoiding traumatic issues in interviews could be seen as disregarding their feelings or experiences (Shaw, 2011).

Despite opposing perspectives on this, scholars appear to agree that researchers need access to resources to manage such scenarios as they commonly occur. Corbin and Morse (2003) have discussed the importance of participants and researchers’ having time to settle down and achieve a “readiness” after one-off qualitative interviews, before they leave each other. The value of this is illustrated in the example with Ruby. However, in a few interviews in this study it was not possible to provide this level of support at the end of the interviews as they were either interrupted, usually by a relative of the woman, or ended abruptly due the time we had allocated for the interview having elapsed. Overall, many of the examples discussed here are common relational ethical issues that researchers navigate in practice. Managing relationships and supporting participants takes times and is done while managing other responsibilities.

Trying to Stabilize Inequalities in Research Relationships

Early feminist work on inequalities in the research relationship suggested that when women interview women, empathy and reciprocity, grounded in their shared experiences of being women or their “insider status,” can help to reduce the asymmetry between them (Oakley, 1981). It is now widely considered that though their similar gender and experiences may connect female participants and researchers, power is always present in research relationships (Doucet & Mauthner, 2008). Some authors have described how, by building rapport and good relationships, researchers who draw on feminist methodologies could be seen to be abusing their power by “faking friendships” to encourage participants to open up (Duncombe & Jessop, 2012; Letherby, 2003). Other authors note that however good the relationship is between researchers and participants, researchers will generally gain the most from the research process (O’Shaughnessy & Krogman, 2012). During the fieldwork for this study, I was aware I would benefit professionally from the interviews. In this section I consider two ways I tried to stabilize inequalities in the research by using reciprocity during interviews and when returning transcripts to participants. The following examples illustrate some of the complexities of inequalities in research relationships.

Reciprocity in the interview interaction

During recruitment and data collection, different aspects of my own identity made me both relate to and notice the dissimilarities between myself and the women I met and interviewed. My role as a mother was one aspect of my identity which I drew on in the recruitment in the community groups. It gave me something to discuss at the community group where other aspects of my identity such as being middle-class, middle-aged, and my role as a student or professional from the university, made me different. However, something I had in common with all the women that I interviewed in the study was that I also drink alcohol. Prior to starting fieldwork, I had considered whether I would discuss my drinking practices and thought that unless women directly asked me about this aspect of my life, I would not talk about it. As I only had limited time for the interviews, and I felt that discussing my own drinking practices could distract from the focus on them, I prioritized discussing their experiences. However, when I started the interviews, I noticed quickly that some women seemed to be interested in my drinking practices. This was perhaps because they thought that I perceived drinking as a problem because I was conducting research pertaining to it. I had tried to explain at the recruitment stage that I was interested in understanding what drinking alcohol meant to them and I did not see it as a problem. Nonetheless, as Warin and Gunson (2013) noted in their study about obesity, just studying health practices can construct it as a problem, which was arguably the case here. I was worried that the women would feel judged, and to avoid this, I wondered whether sharing my experiences of drinking could ease this concern.

After the early interviews, I started to bring my own experiences of drinking into conversations where it felt relevant. This is illustrated through my interview with Denise, a middle-class woman, who did not have children and was a similar age to me:

Denise:

…it started when I was 14, which seems quite young, em, well it is quite young isn’t it, 14…really but it’s probably normal, isn’t it?

Katherine:

I think I probably had my first drink when I was around 14 so yeah, that’s probably…

Denise:

Yeah, yeah. Em, and I was em…it was a boyfriend that I’d kind of…I’d hooked up with, em, and he and his friend used to drink cider…

Bringing in the similarities in my own personal experience felt like the right thing to do because to have left Denise questioning whether drinking at the age of 14 was something other people did, had the potential to cause emotional harm. This reciprocal exchange also encouraged Denise to begin to talk more openly which helped draw out a more detailed account of her experiences. In making these disclosures about my own drinking I felt a responsibility to some of the women I met when I was conducting the fieldwork who were experiencing the effects of very heavy drinking. Yet, I needed to respond to Denise’s individual circumstance to make her feel comfortable.

However, I did not use my experience of drinking in all the interviews because my drinking practices were sometimes very different from those of the women I was interviewing. Other researchers have drawn on intersectionality theory to illustrate that structural differences between researchers and participants such as their social class, ethnicity, and disability can make experiences, and the tools they have to engage in reciprocity, very different in interview interactions (Couture et al., 2012; Ochieng, 2010; Skeggs, 2002). In this study, I noticed that some working-class women spoke about drinking at home with children in a different way to me. I realized that this had the potential to change the interview dynamic and this is illustrated by the following extract from the interview with Toni, a working-class mother a few years younger than me. She said that she did not drink at home with her children in the house because she would be putting them at risk:

Katherine:

Do you drink? Would you ever drink in the house?

Toni:

No, I don’t drink in the house.

Katherine:

No?

Toni:

No, even when the kids aren’t there I don’t, I don’t like to drink alcohol in me house, so I don’t know why but (laughter) [Yeah]. Yeah.

Katherine:

I think there’s people that do and there’s people that don’t…

As I do drink at home and have young children, I did not disclose my drinking practices in this interaction. Exploring differences between us and trying to understand them could have been an interesting diversion to take in the interview, however, at the time, my concern was making Toni feel at ease and I felt that mentioning my own drinking would damage any rapport that we were starting to build. Overall, this relational aspect of the power dynamic was not something that could be resolved but needed to be recognized (Pollock, 2012).

Returning transcripts to reduce inequalities?

In the procedural review form, I had proposed to offer to send transcripts back to the women to review after their interviews. The aim of this was to try to reduce asymmetry between myself and participants so they did not feel like I was walking away with their data (Rager, 2005). Some authors (e.g. Forbat & Henderson, 2005; Mero-Jaffe, 2011) have reported the sensitive nature of returning interview transcripts, however, at the time, I was not fully aware of these concerns. My use of this method resulted in one of the most memorable “ethical speedbumps” in the study, where I felt I had caused a participant emotional distress.

The study consent form provided before the interviews gave women the option of whether they would like to see their transcript after it had been typed up and anonymized. In the early interviews, I did not discuss returning the transcripts to them as we worked through the informed consent process, other than to say I would be in touch with them to know which secure methods they would like me to use to send their interview transcripts so that anonymity and confidentiality were ensured. When I proceeded to send them the transcripts, I asked the women to get back in touch if they had any comments or wanted to change anything. A few replied to acknowledge the receipt of the transcript, however, I did not hear back from the first few women I returned transcripts to. As such, I assumed that they were happy with the transcript and did not want any changes to be made. By this point in the fieldwork I did not think to question this as I was still managing the different components of the study, which were quite labor intensive.

However, about a third of the way through the fieldwork, I became aware of the distress that viewing her transcript had caused Hannah, who had a particularly in-depth narrative of several years of heavy drinking. Her interview had been emotional and ended suddenly when interrupted by a member of her family. As discussed above, because it ended quickly, I had not had the opportunity to spend time with Hannah after our interview. Nevertheless, after checking how she wanted me to send the transcript, I had sent it to her via email. She emailed me back a week later to say that she had found reading the transcript upsetting and that she wanted assurances regarding how it would be used in the study. Knowing that this part of the research had distressed Hannah and that I had not adequately protected her from this emotional harm was upsetting and worrying. My lack of experience meant I had not considered how stark her account would look on paper and how difficult it could be to read.

I replied to Hannah in an email explaining how her interview data would be used and offered the opportunity to remove her data from the study. I also explained that if she decided to remain in the study, I would do everything I could to ensure her anonymity. In response to my email, she replied to say that she felt reassured by what I had told her and that she would “be brave” and take part. I was pleased that my email had comforted her and that I did not have to remove her from the study as her interview had valuable data that I felt would be essential for the analysis I was working on. After this incident, following a conversation with a colleague and my qualitative supervisor, I started to caution women about reviewing their transcripts. As a result, fewer participants chose the option to review their transcripts. I provided a disclaimer for those who did want to see their transcript by explaining that it could be upsetting to read, and suggested they had someone with them while they were reading it. This example again indicates the need to engage with the context of the women’s individual lives and circumstances as, while I had thought this method could reduce inequalities in the research relationships, in actuality, it had upset a participant. It also highlights how my lack of experience of using this method meant I was not prepared for the concerns it could raise.

Discussion

In the examples discussed here, I have illustrated how negotiating ethics in practice and working out how best to support participants in semi-structured interview studies occurs alongside other duties and obligations. The examples also demonstrate that engaging with participants’ lives and circumstances requires time and opportunities for reflection. My examples signal that for ECRs, having good supervisory support from an experienced qualitative researcher is extremely valuable. Further to this, using my study as an example of the types of ethical issues that might need to be navigated during qualitative interview studies and the resources that can support this, I want to consider some structural aspects of the contemporary academic health research context that may limit time opportunities for reflexivity and caring for participants (Edwards & Mauthner, 2012; Wiles & Boddy, 2013).

To support researchers when managing relational ethics in practice, some authors have suggested that there should be ongoing opportunities during projects and within research teams to: discuss ethical concerns, make changes to procedures, and to enable researchers to develop their ethical competencies (Attuyer et al., 2020; Guillemin & Gillam, 2004). Nevertheless, this space and support cannot be guaranteed (Dickson-Swift et al., 2009). A challenge for PhD researchers who work on one-off interview studies is that they often work alone rather than as part of a team. Though my supervisor offered me support, I knew she was busy, and as such, I only contacted her about “ethical speedbumps” where I was concerned about a participant’s wellbeing. With the benefit of hindsight, I can note that there were times in the study where it would have been valuable for me to stop, reflect, and introduce new procedures. For example, although I contacted all the women after their interviews via a “thank you” email or letter, it would have been good practice to also call them on the phone to check that they were doing alright after the interviews (Corbin & Morse, 2003). As ECRs in other disciplines have described (Caretta et al., 2018), I also felt a level of pressure to complete the fieldwork. While I kept a fieldwork diary throughout the project, I only stopped and reflected carefully after the fieldwork was finished.

However, the examples referred to in this article do illustrate that I benefited from having a qualitative supervisor who I felt about to go to for advice and comfortable receiving guidance from about relational ethical issues when needed. Yet, I am concerned that the contemporary health research framework and the move toward working in multi-disciplinary teams may limit opportunities for this mentoring or support (Kontos & Grigorovich, 2018). PhD students are now often supervised across disciplines, and postdoctoral researchers can commonly be the sole qualitative researcher within a multi-disciplinary team (Kontos & Grigorovich, 2018). In these types of multi-disciplinary teams the relational ethical issues, which a team that includes experienced qualitative researchers might begin to anticipate, and can prepare junior colleagues for, may not be foreseen. ECRs may also be concerned about disclosing ethical concerns or aspects of the research that may not have been favorable because they might not want to appear weak or inadequate (Batty, 2018; Johnson & Clarke, 2003). Mauthner and Doucet (2008) argue that whereas the traditional role of the research team was to enable more junior researchers to develop their qualitative research skills and ethical competencies, in multi-disciplinary team-based studies, these opportunities may no longer exist. Thus, as the move towards multi-disciplinary research continues, the support available for junior qualitative researchers will need to be considered.

Another related challenge posed by the current health research context, is the preference for mixed method studies, where qualitative research alone is often not seen as sufficient (Kontos & Grigorovich, 2018; Webster et al., 2019). My study was a predominantly qualitative study, but a systematic review was incorporated into the research proposal alongside the qualitative interviews to secure the funding. This systematic review took longer than anticipated at the start of the study, delayed the study, and added an extra level of pressure when conducting fieldwork. Thus, arguably, working as a qualitative health researcher on mixed-methods studies can make it more challenging to solely prioritize the relational aspects of the fieldwork. My examples indicate how I negotiated differing work responsibilities and how these various external pressures could reduce the opportunities available to retain the space to engage with relational ethical concerns.

The issues regarding the health research context are part of more complex debates about the contemporary academia (e.g. Gill, 2009; Rogers, 2017). My aim in raising some concerns in this final section has been to challenge the notion that ECRs can easily find support and respond to ethics in practice in their research teams, and to show some structural issues that may be limiting their capacity to respond in the field. Other authors argue that the best protection for participants are well-trained researchers who are supported by experienced supervisors (Banks, 2018; Corbin & Morse, 2003). However, when researchers and their supervisors, and managers, work in environments where visible research outputs are valued and the intense relational work involved in fieldwork can be invisible, it may make it more challenging for researchers to always give the best care possible.

Conclusion

All ECRs will need to engage in hidden relational ethical work in qualitative interview studies. This includes managing the complexities of participants’ lives and their own concerns alongside those of the participants. ECRs capacity to respond to these issues is likely to depend on several factors including: their own experiences and training, their informal and formal support structures, and wider structural arrangements.

Any recommendations on providing better support to ECRs would ideally focus on making structural changes to the research context which is unlikely to happen immediately. In the last few years in the UK, some health research funding providers have acknowledged the need to improve the demanding health research culture (Wellcome Trust, 2020). In the short-term, researchers need to be better prepared for ethics in practice through relevant training (Banks, 2018). Those who fund research, and academics who supervise or manage ECRs must take responsibility to ensure junior researchers are adequately supported. Ultimately, I hope this reflection on my experiences help ECRs be better prepared for the aspects of fieldwork they will often have to navigate alone. In turn, this will have the effect of benefiting both them and their research participants.