Ethically important moments as data: reflections from ethnographic fieldwork in prisons

Introduction

The concept of ethically important moments, developed by Guillemin and Gillam (2004), has been used by researchers in discussions about appropriate ethical behaviour in research on a diverse range of subjects (see Bowtell et al., 2013; Phelan and Kinsella, 2013; Clark and Sharf, 2007; Ellis, 2007). It provides a useful framework for researchers when considering their role in responding to unexpected ethical challenges that arise from interactions with research participants. Guillemin and Gillam argue for an understanding of ethically appropriate research, which is nuanced by the realities of research fieldwork and as such able to respond to situations that may not have been envisaged at the time of research ethics committee approval.

This article will develop the concept of ethically important moments further by suggesting they are disruptions from the expected course of data collection, which provide opportunities to look afresh at the setting being studied. As such, they have the potential to be a source of data, which differs from and enhances other fieldwork data and adds to understanding by placing the researcher in a different relationship to the field. In trying to find an ethically appropriate response to the moment, the researcher is no longer a bystander or a recipient of information but an actor, seeing the assumptions and values of other actors in ways that might not otherwise be revealed.

A number of examples of ethically important moments from a recent study of deaths from natural causes in prisons in England and Wales will be considered, some of which result from a dissonance between ‘paper ethics’ and ‘real-world research’ (Armstrong et al., 2014). In this study, ethically important moments also arose from the dichotomy between the ethical values of the prison setting and those of academic research. Identifying these highlighted the ways in which the prison setting shapes the experiences of people as potential research participants and reveals the lived realities of prison regimes, culture and relationships. How these questions are resolved is important in terms of ensuring ethically appropriate research. However, the fact that it is these ethically important moments that occurred, rather than others, is significant. In this way, ethically important moments become tools for the qualitative researcher.

Ethically important moments

In the course of a research project examining women’s experiences of heart disease, Guillemin and Gillam (2004) received an unexpected and disturbing disclosure by a participant. The information shared was unrelated to the research subject, but the interaction was not unusual, they argue, when conducting qualitative research. In considering how to respond to a disclosure of sexual abuse, they were led to a number of ethical questions beyond those typically addressed by research ethics committees. For them, the incident highlighted the importance of reflexivity and the requirement sometimes to make decisions that are difficult without prior consideration or an ongoing awareness of the impact of the researcher’s responses may have. They refer to such encounters as ethically important moments, defined as ‘the difficult, often subtle and usually unpredictable situations that arise in the practice of doing research’ (p. 262). They argue that unanticipated events are not necessarily ethical ‘dilemmas’ to be solved, in that they may not present choices between options for action. The best response may be clear, although ‘the moment of response is an ethically important moment for there is the possibility that a wrong could be done’ (p. 265).

Bowtell et al. (2013) further suggest that ethically important moments can be ‘both procedural and emotional’ (p. 652). Procedural ethically important moments challenge how best to comply with policies and procedures intended to produce research of a high ethical standard. Emotional moments that are ethically important are those where some feeling, often of personal discomfort, is evoked. Examples of both will be considered in the case study discussed below.

‘Procedural ethics’ and ‘ethics in practice’

The process of gaining approval for research from research ethics committees aims to anticipate likely ethical challenges. It is, however, widely accepted that unpredicted ethical issues will arise during data collection. The distinction is often made between ‘paper ethics’ and ‘real world research’ (Armstrong et al., 2014) or ‘procedural ethics’ and ‘ethics in practice’ (Guillemin and Gillam, 2004). Researchers will need to make decisions in the field, taking account of the context, because research ethics committees lack the necessary understanding of contextual detail (Boden et al., 2009; Hammersley, 2007). There is an inevitability about these challenges. Katz (2006) contends that researchers ‘often encounter people and behaviour they cannot anticipate’ (p. 500). Furthermore, Hammersley (2015) argues: ‘doing social research of any kind, but especially perhaps ethnography, is almost inevitably occasionally ethically and politically uncomfortable, at the very least’ (p. 33).

Taking Guillemin and Gillam’s (2004) concept of ethically important moments, Murphy and Dingwall (2007) argue that when risks to those being studied arise, they are ‘likely to be served less well by procedural ethics embodied in regulations requiring fully informed advance consent than by researcher’s conscientious commitment to avoiding exploitative and potentially harmful relationships’ (p. 2229). Guillemin and Gillam (2004) refer to this as the researcher’s ‘ethical competence’ (p. 269), meaning the researcher’s willingness to acknowledge the ethical dimension of the research process and their ability to recognise and respond appropriately when ethical issues arise. In many ways, this article tracks attempts to exercise ‘ethical competence’, an attribute that researchers should seek to continuously develop, along with their other necessary research skills. Other prison researchers have experienced similar things. Crewe (2009) needed to respond to prisoners who found his ethnographic methods intrusive. Likewise, Bennett (2016) struggled to know whether he should intervene in a situation that caused him concern in the prison he was studying. Similar concerns will be equally familiar to researchers in other settings. The setting of a prison serves to heighten the vulnerability of research participants, further exacerbated in this case by the sensitivity of the research subject, but as Guillemin and Gillam (2004) experienced, ethically important moments can arise when least expected.

Background to the case studies

The potential of ethically important moments will be illustrated using examples from a study employing ethnographic methods to explore how the growing number of deaths from natural causes in prisons in England and Wales impacts on prison regimes, culture and relationships. The study, funded by the ESRC, considered what factors determine how prison regimes and personnel respond to dying prisoners and to deaths from natural causes. Using a mixture of participant observation, semi-structured interviews and documentary analysis 12 months of fieldwork was conducted in two prisons for men in the north of England. One of these prisons was in the High Security Estate, accommodating men sentenced for the most serious crimes and often serving very lengthy sentences. It is from this prison that most of the examples discussed below originate. The other was a ‘Category B’ local prison, housing medium-risk prisoners and those on remand. Ethical approval was granted by the NHS Health Research Authority, Her Majesty’s Prison and Probation Service and the University of York.

Vulnerable participants and sensitive topics

Given the distress that can be experienced by some people when reminded of death and dying, plans for the fieldwork placed considerable emphasis on minimising harm and providing support to participants who might be upset. Some ethically important moments could be expected, such as a participant becoming upset when talking about a death, but when this might occur was uncertain. In one case, a prisoner who had been an informal carer for a terminally ill peer approached me to tell me the man had died that morning. As he spoke, it was clear he was struggling not to cry. My instinctive response was to prioritise his well-being over any data collection, and to reassure him I was available to listen if he wanted to talk, and would not be gathering data.

However, other moments arose during the fieldwork, which had not been envisaged but which required ethically appropriate responses. Four groups of examples will be examined in depth. Two of the examples that are discussed below are procedural ethically important moments resulting from participants’ reluctance to accept a copy of their consent form and the difficulties of maintaining anonymity in a closed and supervised community. These ethically important moments were each about the ethical obligations a researcher has towards the research participants. Two ethically important moments that evoked emotional responses in the researcher will also be considered, linked to the vulnerability of participants and to the sensitivity of the topic. Firstly, there was a need to respond to approaches from prisoners wishing to participate in the research, but where there were concerns about their vulnerability. Secondly, decisions had to be made about the extent to involve terminally ill prisoners directly in the study which provoked concerns about the vulnerability of the potential participants.

Responding to approaches from prisoners to participate

There was an expectation from one research ethics committee that individuals who were likely to find the topic upsetting could be identified in advance. Consequently each prison’s Safer Custody Team was asked to identify such people to exclude them from the study. In practice, this proved impractical. It became clear that the Safer Custody Team could not know of every case where a prisoner might find the topic upsetting and that to ask them about the vulnerability or otherwise of a particular individual would potentially cause another harm by breaching a potential participant’s anonymity.

There were many occasions when participants approached me and started discussing the research subject, without any opportunity to seek advice from the Safer Custody Team in advance. In the high-security prison especially, the prisoners who approached me were often aware of the research and had been looking for a moment to talk, having prepared what they wished to say. In other cases, the conversations arose from a prisoner simply asking who I was and why I was there. When approached by a prisoner in this way, I needed to decide immediately which course of action would cause least harm: to continue to listen to the prisoner, or to delay the conversation until checks had been made with the Safer Custody Team.

Two examples stand out. On one occasion, a prisoner who seemed well-informed about the research came to sit with me by the entrance to his wing. He was keen to help and told me various things about how the prison responded after a death. He said, matter-of-factly, that there was a prisoner from the wing in the healthcare centre. This then sparked reminiscence about caring for his mother, who had died of a similar condition. His tone was sadder and his voice less confident, and after he’d finished he repeated what he’d said (Fieldnote, 20 September 2017). On another day, a prisoner I’d not met before approached and asked if I was doing the research about deaths in custody. He proceeded to tell me about two men he had known in the prison who committed suicide, one of whom was a friend, and about another man who he’d been helping, who had died of natural causes in the healthcare centre. He spoke quickly and quietly, but there were strong emotions in his voice and he talked about the need to hide emotions in prison (Fieldnote, 19 December 2017).

With both of these prisoners, this was the start of an ongoing research relationship. I chatted with them many times over the subsequent months, with them seeking me out to share the latest news from the prison. But in the first moment that they started the conversation, I needed to decide how to respond in an ethically appropriate manner. Delaying the conversation implied a lack of respect for the prisoner’s autonomy and their ability to decide for themselves. As Moore and Wahidin (2017) state, ‘there are immense challenges in guaranteeing the dignity of research participants within total institutions such as prisons’ (p. 66). I didn’t want to compound the loss of dignity that comes with imprisonment by suggesting these men could not decide for themselves whether to participate. Furthermore, harm to the participant might arise if no guarantees could be made as to when the conversation would be resumed. Instead, I was careful not to direct the conversation to places the prisoner might not wish to explore and to listen empathetically. The conversation was only regarded as data if I was later able to secure their informed consent and was certain that doing so would not be exploitative.

Research involving the terminally ill

The sensitivity of the research topic led to occasions when I felt uncomfortable conducting fieldwork directly involving prisoners known to be seriously ill or approaching the end of life. This was especially the case around one prisoner who was terminally ill and who subsequently died during the fieldwork period. How my fieldwork might relate to his circumstances required much careful consideration:

I’ve been reflecting on what my role as a researcher is around Eddie, who is known to be dying, and who is currently on the wing, I’m told at his request. I’ve felt there have been expectations from nursing and discipline staff in the Healthcare Centre that he would be the focus of my research for a while. He’s been mentioned by prisoners ever since I started the fieldwork, at around the time he was transferred to the healthcare centre. I’ve been concerned not to burden him, not to expect anything of him given he is close to the end of his life. I’ve also felt uncomfortable that my presence and interest would remind him of his impending mortality. And I’ve been worried that I’d been seen by him, or by staff or prisoners on the wing as some sort of ‘data vulture’ circling around for my own purposes, expecting him and them to offer up ‘data’ for my research without being sensitive to the situation they all find themselves in following his diagnosis. (Research Journal, 22 November 2017)

In another research journal entry, it is noted that ‘there is a certain “got one for you” attitude amongst the staff who know about my research; I’m being pushed towards the man’ (Research Journal, 21 November 2017). Coming at a relatively early stage in the fieldwork, these instances presented challenges. Rapport was still being established with the ‘gatekeepers’ and I was aware that my response could impact on the credibility of the fieldwork. I was still, as many prison researchers, needing to manage the impression I was creating and to perform my chosen identity as researcher (Drake and Harvey, 2014).

The literature on the ethical appropriateness of qualitative research with the terminally ill is inconclusive. De Raeve (1994) suggests that there is perhaps no justification for research involving the dying, given the potential for disrespect of the emotional and physical state of people who are approaching death. In contrast, Lawton (2001) argues it is paternalistic to suggest that people receiving palliative care are too fragile to participate in qualitative research and that this population can find a therapeutic benefit in having a voice. Similarly, Barnett (2001) suggests that while it could be a painful experience for terminally ill people being interviewed, participants were glad to have done it. Barnett (2001) also argues that to exclude them from discussions of the issues that concern them would be paternalistic. Kendall et al. (2007) suggest that ‘some people’s desire to participate in research at the end of life may itself be an example of resistance to social death, an opportunity to be an active and participating citizen again rather than an invalid or patient’ (p. 527). In contrast, Hammersley and Atkinson (2007) say that ethnographers need to avoid creating anxieties and give the example of research involving those who are dying which they say may be judged to be unethical.

If qualitative research is intended to include terminally ill participants, Kendal et al. (2007) advise researchers to proceed as if people do not know they are dying, unless they explicitly acknowledge that they are, and to avoid asking direct questions about death and dying. This was useful advice when I discovered that a senior healthcare professional within the prison had told Eddie about my research and said I would go to see him, assuming that was what I wanted. This was an ethically important moment in that how I responded could potentially distress Eddie, either by involving him or excluding him from the research. Spending time on the wing suggested other ways in which data could be collected without requiring Eddie’s participation and without being a ‘data vulture’. There did, however, seem to be something disingenuous and paternalistic about this. Researching the situation of the dying prisoner without informing him or involving him directly was the mirror image of the staff paternalism that assumed he had no agency with regard to not participating. I was also aware that some harm could result from cancelling the meeting; he was expecting to see me. With Kendal et al.’s (2007) advice in mind, I therefore approached him directly, told him about the research in general terms and was guided by him as to the extent to which he wished to participate.

Informed consent and trust

Informed consent is undeniably a fundamental principle of ethical research but it is not always straightforward. More specifically, in the context of prison research such as this, Jewkes and Wright (2016) find that informed consent as envisaged by research ethics committees ‘is hardly practical in reality – many participants would just back off if an ethics pro-forma were thrust upon them at the outset’ (p. 666). Some challenges around securing written informed consent were therefore to be anticipated, especially given the chosen ethnographic methodology. As discussed above, conversations where informed consent had not been given in advance were only regarded as data if the potential participant subsequently signed a consent form and I was clear that asking them to do so would not be exploitative. However, particular ethically important moments arose around the administration of consent forms.

Having stated to the research ethics committees that I expected to give participants a copy of the signed consent form, only one staff member and one prisoner of the 75 participants would accept a copy. In relation to one group of participants, fieldnotes record:

No one really wants a copy of what they’ve signed; I get the impression that it’s seen as unnecessary. In response to his asking, I tell Ned that he doesn’t need to keep it, and he screws the paper up and throws it away. There’s a slight challenge in how he does this, but I say it’s fine and move on and the exchange is good natured, more a resentment of paperwork than me. (Fieldnote, 11 October 2017)

I felt uncomfortable that the procedure agreed with the research ethics committees was not accepted by the participants and had to think carefully about the ethically appropriate response. Eventually I decided that it was more important that I respected participants’ ability to decide for themselves what offered them the most protection as research participants. I could envision that for prisoners keeping a copy of the consent form meant their involvement in the research could be ‘discovered’ during cell searches, and understood that even for staff there were few places where paperwork could be confidential. When I later asked Ned why he hadn’t wanted a copy of the consent form, he gave several reasons. With some passion, he said there was, ‘that much rubbish under my bed’ and talked about his need to clear it out from time to time. This is understandable in light of the volumetric control on how much prisoners may keep in their cells. When he in return asked me why he might want to keep a copy of the consent form, the suggested scenario of wanting to check back on what he’d signed up to was dismissed; he trusted me.

This echoes with comments from another prisoner who told me that people didn’t take the consent form because they trusted me and made their own decisions based on how I presented and what I said. Hammersley and Atkinson (2007) suggest that participants are often more concerned about what kind of person the researcher is than the research itself. In the case of the prison, Ned identified another surprising reason for trusting me. While many prison researchers are concerned that carrying keys will identify them with the prison authorities and place them apart from the prisoners who are their research participants (Jewkes and Wright, 2016), Ned, warming to the subject of why he didn’t need a copy of the consent form

points at the keys in their pouch on my belt, saying that you don’t just get them without being trustworthy, that he trusts me because I have keys. Having keys means that the prison has checked up on me, and decided I’m trustworthy, and that’s good enough for him. (Fieldnote, 4 April 2018)

This is surprising, but echoes Crewe’s (2009) experience of finding that carrying keys was seen as indicative of being trusted and self-sufficient. Carrying keys is an asset to the researcher, demonstrating that their credibility has been established by the prison authorities, who are seen by prisoners as better placed than them to judge someone’s trustworthiness.

Anonymity and confidentiality in a closed and supervised community

As Jewkes and Wright (2016) note, there are practical reasons why the standard research approaches to anonymity and confidentiality are difficult to maintain in a prison setting. Staff often have to bring prisoners to and from meetings with the researcher. The researcher will typically be visible to staff and other prisoners when interacting with people on the wing. Similarly, Moore and Wahidin (2017) warn that anonymity can be hard to achieve:

The prison can be a difficult environment to find a safe, private space to conduct research and to protect privacy interviews should be conducted out of earshot and, where appropriate, out of eye shot of prison officers (p. 67).

Waldram (2009) is more vehement in his analysis of the problem: ‘The panoptic culture of suspicion permeates every activity. Even “confidential” interviews with participants take place in glass walled rooms’ (p. 4).

In the high-security prison used for the fieldwork, conversations with prisoners took place where the interaction could be overseen, and potentially overhead, by staff and other prisoners. Staff on the wing where I spent most time requested that I kept in view of their office on the ‘2s’ landing. On one occasion when I wanted to talk with a prisoner working on the landing below:

The officer on the 2s was fine with this, but the female officer on the 1s less so – she stood by the open door all the time we were speaking, meaning I felt a certain pressure to keep it short. (Fieldnote, 15 December 2017)

Occasionally pre-arranged conversations with prisoners were held in interview rooms. Each time this happened, prison officers had to be told who I wanted to see, call or fetch the prisoner, and give him a rub-down search before the interview commenced.

Prisoners’ involvement in the research was often known to staff but efforts were sometimes needed to keep the content of discussions confidential. In one instance, Dave, a prisoner who had asked to see me, was categorised as ‘no one-to-one’ meaning that a staff member would have to be present (Fieldnote, 10 January 2018). His status gave him less opportunity for confidentiality than other prisoners and presented an ethically important moment with no neat solution. An officer was willing to sit in the room, but since that would have prevented the discussion being confidential, the suggestion was declined. Instead it was agreed with Dave to speak on the wing landing during association time. This was a public place – interaction could be seen and potentially overheard by staff or other prisoners – but agreed to be a better option than having a staff member present who would hear every word.

Something often overlooked in prison research is the difficulty of protecting the identity of staff participants. When staff members participated in semi-structured interviews, as with prisoners, it was hard to avoid their participation being noticed by colleagues. Offices often had glass walls or were accessed through public spaces and most staff interviewees struggled to find a place to talk where we would not be interrupted. Furthermore, if prison officers who routinely worked in prisoner-facing functions were to be interviewed without disrupting the normal regime, other staff within the prison needed to be told who was participating in the research so they could be replaced. This was a foreseen difficulty. Discussion with the Safer Custody Team in one prison, and at the daily meeting of Residential Custody Managers (RCMs) in the other, led to an arrangement whereby, if the officer to be interviewed consented, I contacted the team who prepared staff rotas to arrange cover for their tasks. Other instances demonstrated the limited extent to which officers actually expected or desired anonymity. For example, I attended a briefing for staff to ask for interview participants, explaining that participation was voluntary and any contribution would be anonymous. Fieldnotes record that during my presentation:

The CM says he’ll do it, and I say ‘well that’s your anonymity out the window!’, but he says he doesn’t care. (Fieldnote, 23 March 2018)

On another occasion, attending a meeting of prisoner representatives at which several prison officers were present, an officer spoke up to volunteer to be interviewed, in front of everyone present. I felt uncomfortable; he clearly didn’t mind.

All potential participants accepted the warning that their confidentiality might have to be breached if they revealed information about a threat to cause harm to themselves or others, a threat to prison security or about illegal activities, malpractice or breaches of prison rules. This was a proviso, required by Her Majesty’s Prison and Probation Service (HMPPS), that seemed familiar to staff and prisoners alike, yet as Israel and Gelsthorpe (2017) point out, it is one in which the researcher demonstrates a privileging of institutional loyalty over the interests of research participants. What was significant was the response that this proviso elicited in some staff. They were unaccustomed to being included in the same stipulation as prisoners, and some of them visibly bristled at this part of the consent form. Despite feeling uncomfortable, I persisted, partly in order to fulfil the commitments to a research ethics committee but also because I wanted to treat staff and prisoners alike in this regard.

Discussion

Each of the ethically important moments outlined above illustrates how ‘ethics in practice’ can be more complex than ‘paper ethics’, and reminds the researcher of the ongoing need to foster ethical competence. In most cases there was not one ethically appropriate response. Furthermore, the ethical issues that arose in this study are not unique to prison research. Similar moments could occur in other settings of qualitative research, particularly where the subject material is sensitive, where power differentials are present or where participants are especially vulnerable. The researcher’s own sense of discomfort or unease is important in identifying when this occurs as an ethically important moment. It highlights that something different to what is desired or expected is going on – something out of the ordinary, a disruption or discontinuity from the ‘normal’, which may be peculiar to that setting, or to that particular circumstance or moment. In this respect, as van Wjingaarden et al. (2018) argue, ‘ethical uneasiness can be seen as an important tool’ (p. 328). Identifying an ethically important moment is significant because it places the researcher in a different relationship to the field. The researcher is trying to find an ethically appropriate response to the moment, and as such is no longer a bystander or a recipient of information but an actor, seeing the assumptions and values of other actors in ways that might not otherwise be revealed.

The uncomfortableness that arises from these ethically important moments can be revealing. In this study, ethically important moments contributed to understandings of the research setting in four key ways. First, they served to highlight that death and dying had become a recognised part of the experience of imprisonment, an acceptable topic of conversation, even with a stranger. Approaches from people wanting to talk about death and dying confounded the expectation of ‘paper ethics’ that these are taboo or sensitive subjects, and illuminated how the culture of the prison differed from what had been expected. A dispassionate consideration of the statistics for deaths from natural causes in prison would have conveyed some data, but it was these ethically uncomfortable approaches to the researcher that lead to a fresh consideration of the place of death in the prison’s culture.

Second, ethically important moments highlight key features of the relationships between prison officers and prisoners. These relationships could be discussed with participants in interviews and ethnographic conversations, but ethically important moments provided the opportunity to see crucial aspects of these relationships enacted. In each case, the researcher needed to take a position, and doing so served to highlight the different position most commonly taken by other actors in the field. The lower status accorded to prisoners by staff is revealed subtly through ethically important moments. It is seen through occasions on which staff’s ‘helpfulness’ towards the researcher potentially undermines the rights or needs of the prisoner. It becomes apparent that prisoners’ autonomy is not highly considered. Even in small ways, such as some staff being bemused by a reminder that their confidentiality cannot be assured if they disclose certain things, the differentiation in status between staff and prisoners is reinforced. In this instance, without the unease experienced by the researcher when asking staff to accept this proviso, this would not have been so apparent. These moments starkly illustrate the power differential between the staff and prisoners as perceived by staff, something that was not voiced by staff participants.

What is also shown in these moments is the extent to which prisoners are dependent on others. They have limited control over their own circumstances and have to trust the prison regime as to who has access to them and information about them. Even when paperwork, such as the consent form, may be assumed to provide some protection, in practice limited storage space and the likelihood of cell searches means that the decision to trust the researcher has to be made at face value.

Lastly, the ethically important moments that arose around participant anonymity illustrate the diminished importance of anonymity in the prison setting for both staff and prisoners. Maintaining anonymity is often regarded as crucial to the conduct of ethically appropriate research. However, by considering the ethically important moments that arise from attempting this, it becomes clear that cultural expectations within the setting, as well as practical reasons, mean that anonymity cannot be fully maintained in the prison, even for some staff. It is often not expected and rarely sought. Talking to a researcher is a public activity: neither prisoners nor to some extent staff can keep their involvement private. The lack of anonymity can result from operational needs, or reflect the core values of the prison, for example officers keeping an eye on the researcher because of their responsibility to maintain safety on the wing. Anonymity is not something participants in such circumstances could be expected to comment on; it is an absence in their lives. It is by experiencing an ethically important moment that the researcher realises the ‘normal’ expectations of anonymity and confidentiality of ‘paper ethics’ do not translate into the real world of the prison.

Because ethically important moments result from the dissonance between the expectations of ethical research practices and the ‘normal’ culture of the research setting they demonstrate the practical implications of the underlying culture, values and assumptions more vividly, and with more emotional resonance for the researcher than might otherwise be apparent. The question arises as to why a specific ethically important moment is occurring, why it is this issue rather than another that presents the moment of ethical peril and it is this which offers the opportunity for insights into the field itself.

Conclusion

Guillemin and Gillam’s (2004) concept of ‘ethically important moments’ helps researchers to consider their role in responding to unexpected challenges when conducting qualitative fieldwork. These moments may be procedural, highlighting a dissonance between ‘paper ethics’ and ‘real world research’ (Armstrong et al., 2014) or emotional (Bowtell et al., 2013). They require the researcher to exercise ‘ethical competence’ (Guillemin and Gillam, 2004) in order to avoid harm to participants. Reflexivity is undoubtedly needed to understand what lies behind ethically important moments, but Guillemin and Gillam have also provided the makings of a further tool for field researchers.

Highlighting examples of ethically important moments from one research project shows how these moments can contribute to an understanding of the research setting, as well as the research process. Ethically important moments are thus useful for what they show differs about the research setting as opposed to what was expected. The dissonance between ‘paper ethics’ and ‘real world ethics’ has long been acknowledged as a potential source of data, but ethically important moments reveal a particular kind of data that can be difficult to acquire from other sources. By highlighting a dissonance between what is expected behaviour and what is happening in practice, the culture, assumptions and values of the field setting that have created the sense of ethical dissonance become more apparent. Aspects that participants may not have thought to mention in interviews or in ethnographic conversations are acted out in front of the researcher, with the researcher as a participant in the scenario, trying to discern the most ethically appropriate course of action and witnessing firsthand some of the lived experience of the particular setting. Few other sources of data offer such an opportunity to illuminate the information collected through fieldwork.