The wrongs of protection: Balancing protection and participation in research with marginalised young people

The project

Similarly to Billett (2012), my path to researching ethical issues came from researching an area with many ethical questions. The research project which initiated my line of ethical inquiry was exploring young people’s pathways into problematic substance use. This came about after having worked as a youth alcohol and other drug (AOD) worker for several years and frequently lamenting the absence of a sound sociological understanding of young people’s pathways into problematic substance use. It was of concern that policy was being based on risk factors (i.e. ‘Early age of initiation is correlated with later problem drug use’), rather than broader contextual factors (i.e. ‘What environment is a young person in where drugs are both available and appealing at age 10?’). Research looking at the complex interplay of structure and agency was necessary to create a rich understanding of young people’s experiences so as to best inform prevention and early intervention initiatives.

The project I undertook to do this adopted a mixed-method design and collaborated with two youth AOD services that operate across 17 sites in Victoria. The most significant data source was in-depth interviews with 61 young people accessing these services (15–25 years; 57% male). Interviews sought an oral history of participants’ lives from early childhood to their current situation. Among other things, young people spoke about how it felt to sleep on the streets; how they raised enough money to get by; why cutting their own flesh made them feel better; how being sexually abused affected their intimate relationships; and how it felt to be 12, and kicked out of home because, ‘she got a bit sick of me being, just, um, [sick of] having a daughter’ (Lisa, 18 years).

These young people spoke to me about sex work, neglect, and abuse so severe that I am not sure I will ever sleep well again. They told me how they coped in a world that had treated them so poorly that I was frequently left wondering how it was they were still alive. I visited young women in public housing that was so unsafe I thought perhaps they would fare better on the street; the security door of a teenage girl’s unit had provided little security when a neighbour hacked at it with a machete while she hid in the bathroom with her 2-year-old son. Its shredded remains were what welcomed me.

Unsurprisingly, these interviews were not a joyous experience for anyone. Many of the young men and women whom I spoke to cried throughout the interview. I reminded all of them that they were able to stop or take a break; however, without exception, they were all adamant that they wished to continue. I made sure that each of them knew that I was comfortable with their tears and that they could keep on crying if letting it out was helpful; this was provided I was comfortable that the tears were cathartic and not distressing. ¹ At the end of each interview, I asked my participants what name they would like to be called in my reports. I iterated that they were more than welcome to adopt a pseudonym, yet the vast majority wanted to attach their own names to their stories. I offered participants the option of using their own names as a way of destabilising the stigma that was so frequently attached to their lives; the assumption that their story was not something that they would want to affix their name to is both presumptuous and offensive. It incorrectly places the shame of their story onto them, rather than onto the perpetrators of the neglect and abuse.

While I do not regret providing participants the option of owning their stories, I have thought a lot about whether they would be so open to attaching their names to these experiences were they older, or working in politics, or had something – anything – to lose. I felt the heavy responsibility I have to protect my participants from feeling regretful, while also weighing how much right I had to abrogate their liberty to choose. It was of critical importance to me that my participants’ safety and well-being was not ever compromised and it was made explicit to participants, prior to their consenting, that if at any time I felt that they, or someone else, was at risk of harm, I would need to do something about this. While there is no legal requirement for researchers to report or act upon such concerns, I feel a moral obligation to do so. This said, with the exception of some referrals to various counselling or drug treatment services, I did not need to intervene. It is important that the researcher develop their own moral parameters prior to beginning the research process so that they can be prepared for how they might handle precarious situations. The myriad of ethical quandaries which may arise in the field means that specific instruction on ‘what is right’ is impossible. Therefore, a well-developed philosophical framework to inform ethical decision-making when out ‘doing research’ equips the researcher to make sound decisions (Daley, 2012).

In my own research, I needed to ensure my participants were protected, but not be so paternalistic that I failed to respect their self-determination. This is a dilemma that HRECs frequently face and this point is recognised in the current National Statement on Ethical Conduct in Human Research (henceforth referred to as the National Statement) which notes:

For ethical review bodies, there can be a profound tension between the obligation on the one hand to give maximum scope to participants’ freedom to accept risk, and on the other to see that research is conducted in a way that is beneficent and minimises harm. (NHMRC et al., 2007: 17)

How researchers and ethics committees negotiate these competing obligations can vary considerably and this tension is the central issue of this article. To contextualise and illustrate the complexities of this, I next discuss some of the competing ideas about the most ethical way to conduct this research project.

Ethical review of research with young people

While the need for formal ethics approval for this project was obvious, dilemmas such as the liberalism vs. paternalism challenge noted above highlight that what constitutes best ethical practice is far from absolute. All approaches to ethical review seek to provide a way to define what the most ethical action is. Barratt et al. (2006) observe that research ethics guidelines are in place to protect and assist the research participants; nonetheless, they are not without limitation. Te Riele and Brooks suggest principles should be understood as a ‘provisional resource’ as they ‘are not straightforward recipes leading to perfect solutions, but they can support a person’s own deliberations in relation to a genuine ethical dilemma in a specific (research) situation’ (2012: 11). This is a similar sentiment to that expressed by Kellehear (1989), who suggests that guidelines should be interpreted as a ‘minimum standard’ rather than a definitive authority on the matter. The current National Statement (NHMRC et al., 2007) cites research merit as its first principle and this is because research design and approval must start with the foundational question: ‘Does this research have merit?’ No other line of ethical query need to be pursued if the answer to this question is no, as research lacking merit is unethical. Providing the research has merit, the final three principles guiding the National Statement pertain to how ethical its design is. Specifically, review seeks to consider whether a research project upholds the principles of justice, beneficence and respect.

Curiously, while both my university and collaborating agencies placed great emphasis on being ‘ethical’, how each defined ethical research practice was not always consistent. Meeting the needs of both committees, when they were sometimes contradictory, was a balancing act that required careful consideration. For instance, one of my collaborating agencies would not approve the project unless participants were to be remunerated for their time, whereas my institutional HREC were concerned that this could be seen as inducement. Remunerating participants is a contested practice. Seymour (2012) provides an overview of some of the debates around this topic, which include whether paying participants creates more or less equity in research participation. The core questions concern whether the poor are less able to participate in unpaid research due to financial constraints (time, petrol, childcare); or perhaps they are more vulnerable to exploitation as they may be more inclined to participate in research for the monetary incentive. Similarly, while there is an increasing practice of not disclosing payment until after completion, Ely and Coleman (2007) have argued that this is deceptive and that central to informed consent is that participants are aware of all risks and benefits. In addition to this, Wiles et al. (2007) comment that using this method as an ethical precaution is somewhat tokenistic as it is highly probable that participants are likely to tell others about the payment anyhow.

As well as participant payments, another issue where the ‘most ethical’ approach was disputed pertained to the location of the interviews. Collaborating agencies felt that if young people were able to suggest a location that they would feel comfortable being interviewed in, this would be a small step in attempting to address the inherent power imbalance between researcher and participants as it gives participants options and choices about the research process. Beals expresses the power imbalance well by describing traditional research methods as positioning ‘researchers as possessors of power and young people as objects of knowledge’ (2012: 35). Numerous authors have argued that it is important to provide participants with choices in the research process as it is empowering to be able to contribute to the process rather than only to the data (Billett, 2012; Tierney and Blumberg Corwin, 2007). Providing participants with an option of having a say in decisions about the research does more than just tell young people that they have rights; it encourages them to enact them.

In my own experience, when assessing the issue of interview locations, attention to enabling participant rights was not the focus of my institutional HREC. The committee expressed concern that allowing young people to suggest interview locations would compromise the researcher’s personal safety and this was an experience shared by Billett (2012), who also wanted to provide participants with choice but was constrained by her reviewing committee. Likewise, Tierney and Blumberg Corwin (2007) faced this competing interpretation of what is ‘right’. These researchers highlighted that the assumption from their Institutional Review Board (the UK equivalent of HREC) was that the researcher themselves did not have adequate capacity to assess whether or not the young person’s suggested location was safe and/or appropriate. It is interesting to note that, in this instance, the perceived risks to the researcher were placed as a higher-order focus than making provisions to support the young participants. If there are real risks to the researcher, then this may be justified. However, given that welfare professionals visit these young people at their homes regularly, it is questionable whether this risk is real, or illustrative of review committees’ assumptions about these young people (Daley, 2009).

In discussion with my HREC prior to submitting my application, there was also debate about whether the experiences these young people had been through had made them more vulnerable or more resilient. Vulnerability and resiliency seemed to be framed as mutually exclusive. Various committee members and collaborating partners were either emphasising protection or participation. My experience was that in some ways these young people were vulnerable and in some ways they were particularly resilient. I needed to ensure adequate protections for their vulnerabilities as well as enable participation in a way that respected their resilience. Because these young people had been through such significant experiences they were obviously very resilient. Nonetheless, sharing these stories can open or exacerbate particular vulnerabilities. Researching humans always entails a level of vulnerability; ethical research must comprise due respect to protecting participants, while also respecting and enabling their right to participate (Mishna et al., 2004). Newman (2005: 23) has suggested that framing young people as ‘reliant and of limited competence’ has contributed to vulnerability, rather than resilience, being the presumed starting point in attempts to understand young people’s needs.

When done with both skill and care, providing someone with a safe space to share their story can be cathartic and therapeutic; however, attempting to do this without adequate thought or training can be distressing and problematic. A well-executed research interview provides participants the opportunity for their life story to be afforded some validity in its own right: the interviewer is not a worker assessing their needs, or a magistrate determining their fate. Sharing sensitive stories may be emotionally heavy, even upsetting, but it should not be assumed that preventing this is more important to participants than being given a place to be heard. Kellehear (1989), who researched 100 people dying of cancer, captures the nuances of the protection/participation tension well when he argues that it is arrogant to assume that the discomfort of talking about painful experiences in a research interview is more distressing than being excluded from attempts to be understood.

The research interview itself has potential to open tender wounds. Given that the interviewer is not in a position to be conducting follow-up, it is necessary that they make considered decisions about whether the interview is providing similar benefits to a ‘single session’ style of counselling, or if the participant is at risk of over-disclosure and likely to regret sharing so much. This was an assessment I made on a case-by-case basis in my own research and the nuances of making this decision and its implications I have discussed elsewhere (Daley, 2012). It is pertinent to acknowledge that only once in the 61 interviews did I feel it in the best interests of the young person to actively prevent them from sharing so much of their story. For the vast majority, whether or not they intended to disclose so much, the interviews were a positive experience. As well as my own perception of it, I had two measures in place for assessing this. Fist, all participants bar one enthusiastically agreed to be contacted in the future to see if they are interested in participating in a follow-up study. Second, at the end of each interview, I undertook a debrief with my participants which gave me a way to ensure that they were feeling okay and gave them an opportunity to reflect on the interview itself. This debrief always revealed a consistent narrative. One young woman captured this common sentiment well:

I haven’t talked like this in ages. But for me, like, whenever I talk about the past sort of stuff, it’s a lot better for me to get it off my chest. ’Cause I love it when people [listen] – when I know they won’t tell anyone, when they won’t go off and blabber – I love that. Yeah, I feel good.

This was similar to a young man who had a very cathartic and tear-filled interview and who emphasised many times in my time at the detox centre he was residing in how grateful he was ‘for letting me download’. I obviously emphasised my own gratitude and privilege that he shared so much with me.

While it is an additional benefit if the interview provides some therapeutic benefit to the participant, the most valuable repayment the researcher can make for those who share their stories is to tell them. As Christensen and Prout so astutely state: ‘The task of the social scientist is to work for the right of people to have a voice and to be heard’ (2002: 483). When collecting people’s stories, it is not enough for the researcher to simply not harm participants, they also need to do their best to do something for them. Researching people’s lives is not about simply framing data or making an argument or a contribution of knowledge, it is about providing means for oppressed groups to convey how they experience their life. Kellehear (1989) has written about his own sense of obligation to ensure that his research was not only completed, but disseminated. People participate in research in the belief that they are to be given voice and the social researcher has an ethical responsibility to provide this, although this is an ethical duty that is rarely articulated (Kellehear, 1989).

For researchers working in contexts where they are not bound by any formal ethical governance (i.e. research done within programs and services), understanding the nuances of ethical conduct in research is even more difficult, given the inability to defer to a committee. While it is unrealistic to expect small agencies to develop a process of formal ethical review, it is perhaps the responsibility of the academy to discuss and publish their own processes and case studies of negotiating ethical decision-making in research for practitioners to draw upon. As is demonstrated in my case study, there are competing ideas about some aspects of ethical research design. The fact that there are differing views does not mean that one set of guidelines was ethical while the other was not; the issue was that what defines ethical practice in social research is not concrete. There is no single ‘right’ way to research the vulnerable; moreover, it is essential to always negotiate each scenario individually (Daley, 2012).

The political act of ethical decision-making

Making assessments about good and harm is complex and nuanced. HRECs are guided by ethical principles, without which assessment would be inconsistent and lack justification. Guidelines based on ethical principles provide a way of reasoning decisions, as well as a transparent and consistent process to ethical review (King and Churchill, 2000). However, principles do not exist independently and the protection/participation tension demonstrates how justice and non-maleficence can compete for priority. Additionally, principles cannot provide instruction on being virtuous. Ethical principles focus on establishing what one should do; when who one should be is just as important. King and Churchill are right in stating that:

Understanding the principles that should guide child and adolescent research is an important task, yet identifying the right principles is only the first step in conducting ethical research with human subjects. (2000: 710)

At the core of ethical human research is an inherent respect for a participant’s humanity. A human life holds intrinsic value and respect for this must be afforded. The value of a human life, and why this must be respected, is perhaps most succinctly explained by sentience. The capacity for a person to think, to feel, to perceive – to be conscious of the world and their place in it – is central to a human’s intrinsic value. At the same time, that a person is a sentient being means that they are able to understand injustice, feel betrayal and perceive abuses of trust. To prevent someone experiencing this, respect is essential.

The nature of research – to find out something new – leaves participants exposed to potential risks. Thus, research with human participants needs to be done in a way which respects a person’s intrinsic value by mitigating risk. This may seem common-sensical when spelt out, but the past provides too many examples to allow us to assume that researchers will uphold this respect (e.g. the Stanford prison experiment; the Tuskegee syphilis study; and the research which was the impetus for ethical guidelines – Nazi medical experimentation). There is an undeniable need for there to be an ethical review of research projects prior to recruitment. However, as there are competing ways of defining what is ethical, review needs to be done systematically and consistently, even in instances where research is being undertaken in contexts in which no formal ethical governance exists (and this review may be done solely by the researcher; a checklist to assist with this is offered at the conclusion of this article).

Having acknowledged why guidelines are used, there is obviously variance in how, precisely, they are interpreted and applied, given that each project is different, as is each researcher and reviewing committee. It is because projects vary so much that they need to be reviewed individually. We cannot simply allow all research that fits into category A to be approved and all research in category B to be rejected without paying attention to the nuances of each project. The time, geography, socio-political context, as well as the competencies of the researcher, all shape just how ethical a research design is. And they should. Humans are complex; research done both by and with them needs to acknowledge this (Fisher, 2009). However, reviewing projects individually (and institutionally) prevents them from being seen as part of a broader research context. Research ethics guidelines shape how research is done, and who it is done – or not done – with, and this has political consequences. Inadvertently, those responsible for writing and enforcing ethics guidelines take on the role of political actors. To illustrate this process, I will continue with young people’s research participation as a case study.

The current National Statement (NHMRC et al., 2007) allows for young people to be researched without overlooking their vulnerability. The requirements for working with youth are listed in Section Four of the National Statement and these are the criteria that researchers must meet and abide by to be approved to research the vulnerable. Te Riele and Brooks (2012) are correct in noting that many researchers are unaware of just how much flexibility these guidelines offer. However, while these guidelines provide for the option of the vulnerable being included without compromising their protection; in practice, meeting these criteria can be cumbersome and create extra work.

This work involves things like obtaining a Working with Children Check or police record check; finding counsellors who are able to debrief participants; and requiring a third party to undertake recruitment (thus limiting the ability to use methods such as snowballing which is often the most effective method when recruiting hard-to-access populations). There are also considerations unique to each project which almost certainly will always arise in one form or another. For instance, in my own research, it was necessary to significantly postpone recruitment from one of the sites due to a spate of youth suicides within the local community, which had obviously left both young people and staff with a heavy sense of grief. It would have been inappropriate for a researcher to be getting involved in the site and interfering in people’s lives in such circumstances. Further to this, making ‘appointments’ with the homeless is a futile task and researchers would do well to recognise that understandings of time differ and that they must adopt the same sense of temporality as their participants to avoid getting caught up in infeasible deadlines. More of the ‘work’ involved in accessing hard-to-reach populations is the formal approvals required. It is essential to spend much time in discussion with collaborating agencies about the research purpose and design. Following this, ethics approvals are required and these applications do take a lot of time. This is because reviewing committees – quite rightly – are diligent in their assessments of research applications which seek to disrupt the lives of the highly vulnerable.

Tight timelines and busy schedules lead researchers to avoid having to do this work. Instead, they design research which prevents it. As a consequence, research into the vulnerable often excludes the highly vulnerable. Billett (2012) explains that ‘young people’ is often only those over 16 or those at school. She explains that the inclusion of only those who are easy to access may not be harmful for those included, but leads to a misrepresentation of young people, or an invisibility of those not included. So although the National Statement (NHMRC et al., 2007) has guidelines in place that aim to balance the need to protect with the need to participate, the highly vulnerable are still less likely to be offered the opportunity for research participation.

The strong focus on protecting research participants overshadows the need to ensure research participation. When protecting participants prevents their participation, this impedes their liberty; but this is justified on the grounds that it is morally superior for there to be no research rather than for there to be harmful research. However, Newman (2005) notes that exclusion may create its own dangers and this concern is shared by Young and Barrett (2001). These authors are concerned about the implications of policies and programs that are designed without having consulted those who will be affected. The argument in favour of strong protection, in spite of young people’s self-determination, holds merit if social research is harmful. If this is not the case, and talking to the vulnerable carries less risk than forecast, then there is an ethical issue with precluding participation as muting a population is wrong as it leaves people – almost always groups who are already marginalised – susceptible to be spoken about, or for, and for their experiences to be constructed for them rather than by them.

What are the risks of interviewing the vulnerable?

Establishing the potential risks of researching a vulnerable group is essential. It is also important to consider how likely this risk is to occur. These considerations need to be weighed against the potential benefits; not specifically benefits to the participant, but to society more generally. The likelihood of these benefits occurring must also be considered. Ideally the equation would see a very low likelihood of risks and a very high likelihood of benefit. Interviewing vulnerable young people about their lives poses three issues which warrant attention:

All of these concerns are valid and researchers should spend time considering each of them thoughtfully. The key issues of informed consent were discussed earlier, though it is useful to reiterate Bessant’s (2006) point, that arguing that young people should have the right to research participation does not imply that they should be treated the same as adults; a position shared by Christensen and Prout (2002). Further, the recommendations from Mishna et al. (2004) to explain to young participants how to enact their rights in research is an example of how young people can be provided the right to participate without abrogating their rights to withdraw consent while also being sensitive to the level of development.

The potential emotional distress to participants is also a legitimate concern. Asking people about sensitive topics and traumatic events must be done with caution, skill and sensitivity. Researchers must make considered assessments about the best interest of the participant; upholding research integrity requires researchers to prioritise participant well-being over their own interests and quest for good data. Participants should be made aware of interview themes prior to consenting to an interview, as well as be made aware that they are able to not answer questions, or that they are able to set parameters for topics they would not like speak about. This leaves young people able to set the terms of their research participation and, with all of this information, young people may choose to participate and to discuss topics which make them sad. It cannot be assumed that sadness or anger implies distress, nor that these emotions are harmful; this said, these stories do need to be received with sensitivity. Beals (2012) is truthful in her observation that the stories that young people want to share may not always be happy and that researchers need to provide a safe space for them to tell these. In allowing young people to share their stories, often accompanied by their tears, the researcher does need to be very sensitive to what is in the best interests of that particular young person. Some young people may be disclosing more than they intended whereas for others the experience may be highly cathartic. The interviewer needs to make these assessments and negotiate the interview accordingly (Daley, 2012). The intention of ethical review should not be to prevent the research from occurring, but to ensure that the researcher is suitably capable of doing this.

Finally, the potential for the researcher to experience a degree of vicarious trauma when researching other people’s traumas is an important issue. There has been little discussion of the topic, although both Kellehear (1989) and Seear and McLean (2008) have discussed this ethical concern as a ‘silent issue’. HRECs often focus on a researcher’s physical safety, but the likelihood that the researcher may experience their own emotional distress is much more probable than the researcher being attacked in the housing estate in which they propose to conduct their research. During the course of my own research, I did at times begin to question my sense of personal safety, but not because of a threat when out in the field. Rather, it was from hearing over and over again the horrific levels of abuse and sexual violence that are a common experience for so many in our community. I felt less safe living in a world that these young people had not ever been safe in. As well as macro-level conceptual ideas about ‘safety’, I also experienced a heavy fog of emotion following intense periods of analysis and writing where the gravity of these young people’s traumas was not softened by their physical presence to remind me that they were okay.

These concerns justify much more attention, and many other authors have discussed, in detail, matters of consent (e.g. Cloke et al., 2000; Carroll-Lind et al., 2006; Renold et al., 2008; Stuart, 2001; Tierney and Blumberg Corwin, 2007). Emotional distress and researchers’ well-being are two under-researched areas that need more work. If it can be established how often, if at all, participants become distressed, HRECs would be better placed to assess the likelihood of risks and benefits (Runyan, 2000). Likewise, if there was more discussion of the emotional risks to the researcher, rather than a primary focus on their physical safety, more attention could be placed on managing the issue in a more systematic way (at present, it is largely left up to the researcher, or research supervisor in the case of student researchers). What also warrants attention is establishing the benefits of research participation. This article has focused on a human rights aspect; young people must be provided an opportunity to have a say in matters which affect them because it is their right. But in addition to this, it would be useful for there to be research exploring the experience of research participation. Rather than assuming what is distressing and assuming that there is little benefit to individual participants, it would be instructive to ask the participants themselves how they feel. I found that conducting a short debrief at the end of a research interview was a useful strategy. It is important to do this because, like young people not invited to participate in research, for those who do participate we have only a constructed picture of their experience of it. Until we ask the people we are talking about how they felt, we are talking for them (Newman, 2005; Young and Barrett, 2001).

Strategies for designing research which protects participants and enables participation

In this article I have used my own research as a case study to illuminate the tensions between protection and participation in research with the vulnerable. Some strategies I adopted to balance these two concerns included recruiting through a welfare service, the benefits of which were two-fold. The first was that it ensured that young participants had advocates. The second was so that I could be reassured that these young people had suitable people to go to were they to feel upset by anything the interview elicited. I wanted to ensure that these young people had the opportunity to participate, but to facilitate that in a such a way that their well-being was not compromised. I also left the task of advertising the interviews with workers who would let clients know that a researcher would be coming in and, if they wished, they could participate in an interview. The workers had no vested interest in whether or not young people participated and this was made clear to the young people. Further, giving young people some autonomy over where the interview took place empowered them as active agents in the research process. This is something that was equally important in building rapport with these young people prior to the interviews and it asked them to be active participants in the research process rather than simply be research ‘subjects’. Additionally, spending time in spaces where young people had a sense of belonging meant that I was getting to know them, and them to know me, in places that they felt were in some sense their territory. Balancing protection and participation in the interview itself required careful thought. While my interviews were semi-structured in nature, I did follow an interview schedule. I had designed this with the young people’s comfort in mind. When collecting life-history material, it makes logical sense to start at the beginning of a person’s life, but I knew that early childhood was likely to be an intensely painful time for many of my participants. Instead of beginning with this, I started with, ‘So tell me about primary school.’ This question was purposely broad so that participants could say as much or as little as they felt comfortable with while we were building rapport. As this rapport developed, I explored more sensitive areas before winding down with a strengths-based conclusion where I asked the young people to tell me how they had managed their adversities, their best qualities and their hopes for the future. This was a way to increase the probability that participants would leave realising their strengths. I finished with a short debrief where I was able to check-in on them, as well as give them a place to have the ‘final word’.

While there is no perfectly ethical way of doing research, the following questions are instructive in helping researchers to ensure that they are thinking broadly about the ethics of their project and would be especially useful to practitioners without access to a reviewing committee.

Conclusion

As the current understanding of ethical research is that it is research which avoids harms, exclusion of some potential participants does not seem, in itself, compelling enough to warrant changes to the review process. While it may not be ideal that some people are unable to participate in research, it is argued that research which is not ideal is preferable to research which may potentially harm the vulnerable. The issue of significance is that the highly vulnerable become excluded from having a say in matters which will affect them and, in a libertarian society, that can only be justified where there are potential harms to participants. In social research, potential harms can be hard to assess. While there is always concern that a participant may become highly distressed when talking about sensitive issues, how often this actually occurs is unknown. The deficiency of research into how research participation affects participants leaves the researcher ill-equipped to make a case that their methods are not risky.

There is often an assumption that vulnerable people may be harmed by talking about their experiences; but another assumption – that being granted the opportunity to share their story may be therapeutic, validating or empowering – is not afforded the same weight. While it is essential that HRECs are cautious in their assessments, it is important that their caution is appropriate and does not cause them to overlook – or create – other issues. There is a very serious harm which arises from not providing the vulnerable with an opportunity to participate. The exclusion of people from research, in the name of ‘protecting’ them, prevents research from being able to give voice to oppressed groups, and thus limits opportunities to advocate for change in these people’s circumstances. Parts of a population can become invisible because they are either so tightly protected or too inconvenient access. It is imperative that consideration is given to the potential wrongs that arise from convenience sampling and the exclusion of especially vulnerable groups. Encouraging research with the marginalised requires more thought, time and effort, in both designing and undertaking research; but it is only with their inclusion that the principle of justice can be truly achieved.