Resolving ethical challenges when researching with minority and vulnerable populations: LGBTIQ victims of violence, harassment and bullying

Despite greater attention being paid by researchers to the practice of researching with minority and vulnerable communities, there is little in the literature that details how to render meaning to theory in practical terms of conducting ethical research with the LGBTIQ community. Drawing from a study conducted at a university in Victoria, Australia, with LGBTIQ university students in order to explore their experiences of harassment, bullying and victimization in tertiary settings, the authors detail the ethical challenges faced, and their responses to these.

The authors sought to gain a contemporary understanding of Australian LGBTIQ students’ experiences of victimization. Literature from American studies suggests that victimization of LGBTIQ youth in university settings remains a significant issue (Jayakumar, 2009; Oswalt and Wyatt, 2011; Tetreault et al., 2013). Research has indicated that for LGBTIQ community members barriers to services exist, creating negative impacts on social and medical well-being. Members of the LGBTIQ community cannot assume services such as education are LGBTIQ safe and friendly or properly tailored for members of their community (Todahl et al., 2009). The authors sought to explore the following: experiences of harassment and violence of LGBTIQ students in the Australian context; the effects of these experiences on LGBTIQ student outcomes; determination of the extent of accessibility of reporting measures; and the notion of ‘safe’ and ‘unsafe’ spaces on campus. A central component of the research was to explore both good and bad experiences, and to allow for the identification of areas where provision and support for LGBTIQ university students could be improved.

A range of ethical issues arose during the conceptualization of this project, the recruitment, interview and reporting stages. The authors reflect on how they responded to these ethical tensions utilizing their theoretical foundations. Below, a brief overview of the project and its conceptualization, and theoretical considerations and methods, is provided. This is followed by five ethical challenges that arose during the development of the project, its enactment and in reporting stages. These comprise the following: (i) unethical research and concern with LGBTIQ populations as the object of academic research; (ii) recruitment of participants and ensuring sufficient participant autonomy without research bias; (iii) participant self-labelling, categorization and maintenance of confidentiality; (iv) responding to future and potential harm to participants; and (v) disclosures of violence and the achievement of justice. The authors reflect on the resolutions to such challenges, alongside their utilization of a feminist and queer approach to research, enabling them to empower participants and avoid conducting research abuse.

Although Renn (2010: 135) suggests that ‘no longer can it be said there is a “gap in the literature” on lesbian, gay and bisexual college identities’, much of the literature focuses on non-Australian tertiary or early-age (primary and high-school) instances of LGBTIQ victimization (Diaz et al., 2010; Chesir-Teran and Hughes, 2009). Jayakumar (2009) suggests that in the United States, LGBTIQ students often encountered a hostile climate around campus, despite many universities generally being heralded as progressive and liberal spaces (Yost and Gilmore, 2011). Research also suggests that LGBTIQ students who had experienced bullying and harassment were not likely to report such incidents (Grossman et al., 2009). The impacts of victimization can be significant, with studies suggesting lower academic outcomes and lower self-esteem (Kosciw et al., 2013; Pearson et al., 2007), and that distressed LGBTIQ youth are more likely to be engaging in risky behaviours, such as alcohol, drug and substance use (Corliss et al., 2010; Marshal et al., 2008; Stall et al., 2001), and be at a disproportionate risk of suicidal thoughts and attempts (Grossman and D’Augelli, 2007; Peter and Taylor, 2014; Scourfield et al., 2008). In addition, evidence suggests that LGBTIQ people frequenting educational environments that are hostile or at least unwelcoming of their gender status and sexual orientation reported elevated levels of unexcused absences (Robinson and Espelage, 2011), feeling unsafe, and actively avoiding places such as locker rooms, bathrooms, cafeterias and grounds (Kosciw et al., 2012; Taylor and Peter, 2011).

Conceptualized as an interdisciplinary study in the disciplines of criminology and sociology, the project was undertaken by researchers, both post-positivist in their approaches, with an acute awareness of the ‘centrality of language in the production of the individual and the social’ (Williams, 2006: 232). While adopting a naturally interpretivist stance, great value was placed on the opportunity to hear and collate participants’ narratives, and it was recognized that meaning is not fixed, nor is it readily apparent. The project utilized interpretative phenomenological analysis (IPA), the process of examining experiences and meaning-making of participants who may have shared experiences or multiple perspectives on a shared experience (Larkin et al., 2006; Reid et al., 2005) – in this case, the shared experience of being an Australian LGBTIQ-identifying undergraduate student. Such studies utilize what is known as a homogenous or purposive sampling technique. IPA often employs a grounded theory approach, where data are analysed in order to develop themes or codes, rather than these being developed and then searched for, within the data collected (Mith, 2011). This qualitative research project was informed by feminist and queer theory approaches during the methodological design and conduct of the research. The recognition and utilization of a ‘feminist ethic of care’ in the approach (see Gilligan, 1995) requires the researchers to reflect on the concept of connection and the connection between people as a fundamental facet of human life. The feminist ethic of care requires the questioning of patriarchal structures of understanding our own relationships and community. In particular, it helped facilitate solutions to, and to avoid what has been termed ‘research abuse’ (Gerrard, 1995). Gerrard (1995: 62) notes that this, defined as the dubious ‘practice of researchers parachuting into peoples’ lives, interfering, raising painful old feelings, and then vanishing’, has the potential to leave ‘the participants to deal with unresolved feelings alone and isolated’ without additional support. This is particularly important in studies such as this one, engaging with a minority population that has been extensively victimized.

Feminist approaches to qualitative and methodological research utilize a ‘feminist orientation’ of understanding along with an ‘ethics of care’ developed from a critique of the social engendering of caring work onto women, and thus its subsequent devaluing and reduction in importance in social, political and cultural life (Edwards and Mauthner, 2012; Preissle and Han, 2012). In such a framework, research ethics is not only conceptualized from a philosophical and moral framework, but also considered in the practical application of relational ethics, such as in the research process (Edwards and Mauthner, 2012; Preissle and Han, 2012). As such, a feminist ethic of care draws from moral obligations concerning attentiveness, responsibility, competence and responsiveness, for example, within the qualitative interview and other research approaches (Tronto, 2005). These latter obligations are understood as central to, and form an ethical responsibility within, the research process, and are expected of the researchers (Gilligan, 2008; Halse and Honey, 2007).

Furthermore, the use of a queer theoretical approach responds to a number of longstanding and emerging issues in conducting appropriate research with LGBTIQ populations. Specifically, there has long been a concern with the tendency to view LGBTIQ citizens as having a disease, and the attempt to pathologize the LGBTIQ community within medical and clinical discourse (Price, 2011). Although more recently there has been an increase in the literature and focus on the LGBTIQ community, this generally privileges white gay male experiences above others (Kong et al., 2002; Price, 2011). Post-modern queer-theory approaches to LGBTIQ research take a ‘queering the field’ stance, whereby a multiplicity of narratives are acknowledged (Kong et al., 2002; Price, 2011) and identity labels are used only as qualitative markers rather than as strict categorical boxes (Beasley et al., 2015; Better and Simula, 2015). Often in queer-theory approaches a mutuality exists between the interviewer and the interviewee (Price, 2011), and the interview itself is conducted as a space in which interviewees are given the opportunity to be reflective, reflexive and self-aware (Kong et al., 2002; Price, 2011). Within this study, the authors ‘reject[ed] the hierarchical relationship between the researcher and researched to instead act in a mutual relationship’ (McClennen, 2003: 35) and space was given to the participants to detail their lived experience in a way that was authentic to them, for example an open-ended rather than closed-category approach was taken in the demographic survey.

To conduct this study, the authors utilized a mixed methods approach comprising a brief demographic survey and semi-structured interviews. The aim of the survey was to collect demographic information for the purpose of knowing what language or terms are being used to denote non-normative lived experiences of sexed bodies, sexuality and gender expression and identity. Interviews were conducted using a progressive-focusing approach – a method that utilizes an interview guide as a loose structural tool in order to allow for the development of organic conversation; open space is provided to explore unexpected discussions and reach data saturation (Fontana and Frey, 2000; Guest et al., 2006; Schutt, 2006; Strauss and Corbin, 1990). To participate, volunteers had to be enrolled as an undergraduate student, over the age of 18, and identify as LGBTIQ, questioning, or non-gender binary. Postgraduate students were not included, to limit the focus of the pilot study. Participants were recruited using physical posters around the campus, advertising in classes and lectures, and online through social media.

Interviews took place with 16 students who fitted eligibility criteria, out of a potential of 26 who indicated an interest in the project (see Table 1). Most students who participated identified their sexed body as female (63%); their gender identity as woman (50%); their gender presentation as feminine (38%); were either 25 (31%) or 20 (25%); had been studying at the university for 2.5 years (31%); and were Anglo-Australian (63%). They studied primarily in the Faculty of Arts (56%), and were under taking either a Bachelor of Arts (38%) or a double degree (44%), with 50% of students stating that sociology was either their primary major, or one of their majors for combined studies. Sexual identity was much more varied, where 19% of students identified as gay, 19% bisexual or 19% queer and 13% pansexual. In the following sections, we detail the ethical challenges faced when conducting this research study, alongside the attempted resolutions to such challenges.

Warner (2004: 335) calls for researchers to ‘uphold [their] end of the bargain’ suggesting that the good faith of participants is ‘built on an implicit belief that by surrendering privacy and exposing themselves to our gaze and categorization system, their life and the lives of people in similar situations will improve’ (Warner, 2004: 335). Examples of unethical research on, and utilizing members of, the LGBTIQ community are numerous. These have included the willing participants who engaged with Krafft-Ebing’s and Chaddock’s (1893) early studies into LGBTIQ sexualities who had no idea they would be characterized offensively, medically pathologized and discussed negatively (Bauer and Wayne, 2005), as well as examples of involuntary participation and coercion (see Humphreys, 1970; Martin and Meezan, 2003). In addition to the harm caused through the operationalization of the research and methods engaged, harm can also be caused through the ends to which the research is employed. Herek (1998: 247) cites studies by Cameron and colleagues (Cameron and Cameron, 1996; Cameron et al., 1985) that have been used to ‘promote stigma and to foster unfounded stereotypes of lesbians and gay men as predatory, dangerous and diseased’. Such research was used to support legislation blocking equality and promoting the continued discrimination of lesbian, gay and bisexual citizens (Herek, 1998).

During the development, construction and implementation of this project the authors sought to be reflexive to the potential (and perceived) impacts of the research on the participants, the wider community, and the researchers themselves. A key component in the conceptualization of the current project was the authors’ desire to ‘never report results without … anticipating and confronting ways which the popular media or public might distort or misinterpret them’ (Martin and Meezan, 2003: 195), in addition to avoiding the supporting or fuelling of stereotypes of LGBTIQ community dysfunction. In thinking through how to contribute to the growing body of literature and in order to have practical impact to prevent and respond to incidents of bullying, harassment and victimization of the LGBTIQ community, it was necessary to give significant thought to how the data could be manipulated and used in a way contrary to the intended aims.

In the process of reporting the findings the authors faced a dilemma. Although selective reporting of particular events could prevent fuelling perceptions of community dysfunction, this may act to censor and fail to accurately represent the data captured. In resolving this issue, the authors opted to report holistically the incidents that occurred, ensuring that data were presented in an honest and representative way. For example, the benefit to community members in reporting the previously under-reported intra-community incidents was sufficient to outweigh potential misuse of the data. In Roffee and Waling (forthcoming 2017) the authors focus on LGBTIQ perpetrated abuse as opposed to that perpetrated by non-LGBTIQ persons. It was noted and emphasized that this was one form of a number of types of victimization of LGBTIQ community members. While the results could be manipulated to further pathologize the LGBTIQ community, appropriate and necessary caveats to the data are necessary to provide balance and to counter potential misuse.

The research was labelled high-risk within the internal university ethics review process because of the subject matter under discussion and the inclusion of two trigger categories. The first was the inclusion of LGBTIQ students (people who would not usually be considered vulnerable, but would be considered vulnerable in the context of this project), and secondly, that the project involves sensitive and contentious issues, for example suicide, eating disorders, body image, trauma and violence. The authors recognized that participation in the project might be difficult for a number of interviewees, in the context of discussing experiences of violence as the result of their sexuality or gender. The researchers were aware that their questions would in some cases require students to recall traumatic incidents that might ‘touch upon painful life events [that] may generate considerable emotion and even distress’ (deMarrais and Tisdale, 2002: 191). Additionally, participants may not have previously considered their sexed, gendered and sexual identity as precipitators of crime committed against them.

During project conceptualization the authors noted a desire capture activity that could broadly be classed as violence, harassment or bullying, ranging from less serious though damaging verbal and non-physical acts, to serious hate-motivated and violent attacks. Warner suggests that ‘often a queer researcher may eschew offering a clear definition of their terms, for they do not want to risk essentializing or reducing any of the categories’ (Warner, 2004: 326). In seeking to recruit students to the project and ensure they maintained participant autonomy, the authors stated they were interested in interviewing LGBTIQ students concerning their experiences of harassment, bullying and/or violence and (as an LGBTIQ student) and opted to use those words. The authors decided that the choice of terms used was sufficiently vague to encapsulate a wide range of experiences and would indicate to potential participants the type of research to which they were agreeing to contribute. A number of potential participants contacted the researchers to express an interest in participating, though they noted that they felt that they did not meet this eligibility criterion using their own personal understanding of their experiences. Although the data collected show that all of the participants experienced, on some level, harassment, bullying and/or violence (to be discussed in a subsequent article), they did not all perceive their experiences as such. When contacting the researchers many of the participants stated that they did not experience violence, and others did not believe their experiences were sufficiently serious, yet both groups wished to volunteer nonetheless. The authors opted to interview the students who wanted to participate as the authors believed that during the interviews there may be discussion of experiences that fit within a broad understanding of harassment, bullying and/or violence even though the participants did not label them as such.

Had these participants been excluded on the basis that they did not feel their experiences equated to violence or harassment, the authors would have (i) missed a chance to identify the ways in which discrimination and violence have shifted on campus, in terms of invisibility and marginalization, thus denoting a changed landscape regarding how violence is understood, and (ii) had limited participation. Avoiding mention of the terms harassment, bullying and/or violence and then discussing such topics within the interviews may have negated the informed consent and resulted in subjects not being informed about the nature of the research. Unfortunately, the use of the words harassment, bullying and/or violence resulted in the participants having prior beliefs and expectations about what they thought the researchers wanted to hear about, and thus this could have prevented them from disclosing the more insidious and normalized experiences of contemporary LGBTIQ life. The researchers are cognizant of the need to utilize, in future, alternative approaches to categories that would satisfy the requirements of the university ethics process and ensure that the research facilitates autonomous participation.

When constructing the project the authors recognized the challenge for researchers posed by the term LGBTIQ. A queer identity ‘demarcates not a positivity but a positionality vis-à-vis the normative – a positionality that is not restricted to lesbians and gay men but it is in fact available to anyone who is or who feels marginalized because of her or his sexual practices’ (Halperin, 1995: 62). The authors were aware of ways in which, through the practice of the research, they could act to exclude community members who do not identify with the LGBTIQ acronym. Brickell (2006) and Chambers (2002) noted that social science approaches have a tendency to name categories to denote a discursive experience, and this naming is unable to capture the complexities of individual or collective experiences. Although language plays an important role in understanding social phenomena, it has the tendency to describe pre-existing entities that, as Brickell (2006: 100) and others (Chambers, 2002; Ussher, 2000) maintain, ‘exist in a neutral state awaiting labelling’. Thus, in attempting to do research with a queer population, researchers are left with the dilemma of how to be inclusive and recruit those who may no longer, or perhaps never, identified with these categories as reflective of their experiences. For example, using phrasing such as ‘non-heterosexual’ continues to privilege heterosexuality as the norm, and is exclusive of gendered experiences, such as those of the trans community. Although diverse genders, sexes and sexualities could be a preferred term, it is not as widely recognized as the acronym LGBTIQ.

Scholars such as Beasley et al. (2015) and Better and Simula (2015) have advocated for new ways of ‘labelling’ or ‘categorizing’ experiences of gender and sexuality when conducting empirical research. Beasley et al. (2015) note the linguistic nightmare of the interpretation of category labels, arguing that concepts such as ‘orientation’ can be too linear and unable to capture the complexity of the individual experience. However, they do not suggest the abandonment of terms. Rather, they advocate for the use of such terms including heterosexual, LGBTIQ or gay as ‘qualitative markers’ to allow researchers a starting point in their empirical investigations (Beasley et al., 2015: 692). Similarly, Better and Simula (2015: 668) note that the use of category identification can erase significant findings and data. As such, there is a call for a research approach that allows individuals to self-define their orientations as a way to continue to capture the complexities of sexual and gendered identities (Better and Simula, 2015: 679). Utilizing a queer perspective, the authors employed this practice through empowering participants to define their selves, rather than having to adhere to a strict set of categories provided for them. The categories of LGBTIQ, questioning and non-gender binary were used in advertising and calls for participation, in an inclusive way in an attempt to capture the variety of queer experiences.

Related to this issue of self-categorization is the concern the authors had with how to balance the issue of confidentiality and potential identification of the research participants when reporting the research findings. The authors were concerned about the possible identification of the participants who took part in the research project. The empowerment of the participants to be able to self-define their identity is in tension with the researchers’ desire to accurately report the findings of the research. With the LGBTIQ population on campus being a small group, there is added concern that the reporting will lead to participant identification. Within a constructivist paradigm the authors were attempting to locate and document lived experience. Although significant thought had been given to the eligibility criteria concerning location of identity (sex/gender/sexuality), it was unclear how this would operationalize until after the data collection stage of the project.

The demographic survey contained open-ended questions to allow flexibility in how students identified themselves. The authors provided descriptors to aid the participants in responding to questions on gender identity, gender presentation, sexual orientation and sexed identity. Through using the approach to identity as qualitative markers rather than categorizing particular experiences, the authors were able to see more clearly how students used terminology to define their identity. However, this has a significant impact on the reporting. The authors were concerned that with such a small number of participants; the diversity of the participation; and the use of some identifiers, the participants may become identifiable. In particular, the authors were concerned about mislabelling or re-labelling the participants’ gendered and sexed identities, particularly because they had empowered the participants to self-identify. A broad range of sexual orientation identifiers were used, including queer and pansexual, as opposed to more traditionally used identifiers as gay, bisexual, or lesbian. Further, additional descriptors, such as bi-curious, agender, androgynous, masculine/feminine occ. androgynous, and less categorical descriptions such as ‘fluctuating/off the scale’ and ‘complex’ were used by participants to describe gender presentation, identity and sexual orientation. The recognition of multiple precipitators for victimization (including race), and the difficult-to-determine, unclear and grey spaces are focal points of interest to feminist and queer research. The identification of the multiple characteristics of identity is thus foundational to the research. However, the concern about identification of participants is amplified with such a small sample and participants coming from a diversity of backgrounds. The authors thus shared a concern surrounding whether a deliberate failure to mention the ethnicity or nationality of the participant, in order to ensure that they remained unidentifiable, was appropriate. In particular, the concern was that the authors would be perpetuating the silencing of non-white and non-stereotypical narratives.

The project as conceptualized was concerned with challenging existing narratives of experience and providing space for the airing of alternative narratives. However, the authors faced a challenge in how to detail those non-stereotypical narratives and ensure confidentiality. Although this type of ethical consideration is not new in narrative research (see Esin et al., 2013) it is not easily resolved and often requires a pragmatic approach to reporting. When discussing incidents involving participants where there were apparent intersecting factors for victimization, often the reporting of the intersectional concern is central to the discussion. In some cases it may not be appropriate to report all demographic data, even when this is relevant. For example, nationalities would be reported when appropriate, and broader categories of reporting would be used, e.g. a participant from Samoa may be described as being from the Pacific region. It is therefore possible to draw attention to the intersectional nature of the victimization without detailing particular characteristics that could be cross-referenced for identification.

By nature of having human participants, the authors were required to comply with the university ethics approval process. All Australian universities require adherence to the Australian Code for the Responsible Conduct of Research 2007 (NHMRC, 2007). While it has been suggested that research ethics is concerned with the protection of participants who may be viewed as vulnerable to coercion or harm (Dalton and McVilly, 2004), it also provides an opportunity for university administrations to co-opt research and to impose or restrict research agendas. There has been a growing body of literature that has analysed the operation of institutional review boards in the USA and university research ethics committees in the UK, which suggests that they are operating to mitigate risk to the institution (Adler and Adler, 2002; Haggerty, 2004; see also Robertson, 2014: 72). Hedgecoe (2015: 10) suggests that ‘it is perhaps unsurprising that research ethics committees become co-opted by university management to prevent research which might embarrass the institution’. In addition, Colnerud (2014) cautions against allowing the university ethics approvals processes to narrow a researcher’s field of vision and their sensitivity to detect ethical risks. The authors were thus careful to structure the project so as to ensure that the research met both the requirements of the university ethics process and avoided becoming a target of (perceived or otherwise) managerialist tendencies to influence the research agenda and remained ethical to participants.

On reflection, being guided by the ethics approval process and the bureaucratic procedures did, as Johnsson et al. cautioned against, and entailed a risk of heteronomy and that the ‘important but less “manageable” moral matters [were] left unaddressed’ (Johnsson et al., 2014: 39). After a number of interviews had been conducted, the authors undertook an interview in which the participant declined to take the explanatory statement, statement of reporting and the list of counselling services on the basis that this would put the participant at risk. Significant thought was given to potential sites of harm or risk, including the location of the research, how participants get in contact and the conduct within the interviews. However, the authors had not given thought to the risk created by a component of the procedure and integral to participation that was required by the ethics committee. The participant in question noted that their ethnicity and cultural background made their participation more hazardous than the authors may have anticipated. They articulated that their participation in the project placed them at risk from other family members. This risk extended to doing things off campus, such as being seen receiving or viewing an email or a website relating to LGBTIQ issues, and in particular being found in possession of materials from the interview.

Following this interview, the authors quickly reflected on the response from the participant and the implications for the research practice. It was decided to alter the method whereby the mandated information would continue to be provided at the start of all interviews, but an offer to dispose of the information sheets on the participant’s behalf would be made should they ‘deem taking the information sheets to be a risk to their safety’. Thought was given to offering to dispose of the information ‘should the participant not want it’; however, it was decided that the selection of wording was particularly important. Participants may indicate that they do not ‘want’ the information; in particular, they might refuse information relating to counselling because of the stigma attached to accessing counselling services. Therefore, students were given the information as required by the ethics committee; however, they were also given the opportunity to leave the information sheets, should they deem that taking the sheets would pose a risk to their safety.

Although not an example of a direct risk of harm, the researchers faced an instance where a potential harm could be minimized. During one interview a student detailed their feelings of alienation, and queried whether there were others like them on campus. Although the participant did not seek names or identifiable details of the other participants, it placed the researcher in a momentary dilemma as to how to respond. This was a case where ‘ethical codes and guidelines for research projects do not have answers to all of the ethical issues that may arise during research’ (Orb et al., 2001: 95). While recognizing that participation was anonymous and not disclosing any identifying information, the researcher reassured the student that there were others who shared similar attributes to them and that they were not the only person to identify as they did. In reconciling the decision to disclose in the course of an interview, the ethic of care supported a decision to act to mitigate a participant’s feelings of lack of belonging. There was an overriding benefit to the participant in knowing that there were others like them on campus, and the disclosure was limited to information that would later become publicly available in reports and publications relating to the research.

As researchers heavily influenced by a feminist paradigm that has a ‘strong political and a moral commitment to reducing social inequality’ (Cancian, 1992: 626), the empowerment of participants was important in the research process. Although participants were empowered to use whatever labels they wanted with regard to their identity, and although the participants may not have identified themselves as victims of violence, harassment or bullying, it was recognized that objectively the authors may have elicited reporting of incidents that were identifiable as crimes. In its dialogue with the authors, and before approval of the project, the university ethics committee noted that there was potential for liability and ramifications concerning disclosure of incidents involving the university. In thinking through how to respond to a request for clarification from the committee concerning how the researchers would deal with any such disclosures from participants, the authors sought to reconcile the principles of justice and autonomy. Although employees of the university, the authors recognized an overriding duty of care towards the research participants, and neither author wanted to contribute towards further mistrust between LGBTIQ populations and the academy. However, there was a concern that the university ethics committee would not approve the research request without something more.

In order to respond, the researchers agreed that a written ‘statement of reporting’ would be provided to the participants that detailed the purpose of the project as a research project, and the reporting obligations of the researchers. The authors were concerned that the project should be seen as independent and not a university-sanctioned exercise in searching for incidents. Yet at the same time the authors did not want the fact that the research may highlight issues concerning university practice and employee conduct to deter the ethics committee from approving the project.

The authors actively informed the participants that participation and disclosure during an interview was for research purposes and that no action would be taken on their behalf to report the incident to any other body. However, the authors felt that simply by reporting the incidents as research findings and without more, this would be akin to generating ‘research abuse’ (Gerrard, 1995). On a number of occasions during the interviews, where students disclosed incidents including cases of bullying, intimate partner violence and sexual assault, the authors offered to support participants should they wish to disclose to official reporting channels and sources of authority. The authors determined that this approach would not constitute research abuse, as the participants would not be left to deal with unresolved feelings, alone and isolated. In one interview, a participant detailed a very serious incident. The participant ‘wanted it to be on record’ but they did not wish to take any action in relation to reporting the incident to the university or police services. The student felt able to discuss this with the researchers and acknowledged the offer of support that the researchers would provide, should they choose to disclose through official reporting channels. This is a practical demonstration and application of the ethic of care, with the authors concerned about ‘a commitment to the development of another’ (Preissle and Han, 2012: 587). Applied here, the role of the researcher within the theoretical framework allowed for the placement and empowerment of the participant at the heart of the actions taken.

Although concerning a different group of vulnerable participants (bereaved family members), Crowther and Lloyd-Williams (2012) suggest that it may also be helpful to think of research and engagement with participants in terms of potential benefits for participants when the therapeutic aspects of the interview are considered. A number of studies indicate that research participants who disclose incidents of abuse report gaining something positive from the experience (Newman et al., 1999), even if they choose not to report to the appropriate authorities. While detailing many of the known risks in research on trauma, Newman and Kaloupek (2004) cite a number of benefits reported by participants after undertaking such research. Such benefits for participants include empowerment, learning and insight, reduction in stigma, an ability to disclose in an accepting setting, altruism, feeling worthwhile for participation, receiving favourable attention, and knowledge of a kinship with others (Newman and Kaloupek, 2004). Other benefits of qualitative interviews include catharsis, self-awareness, healing and providing a voice for the disenfranchised (Hutchinson et al., 1994), alongside the recognition that there can be significant risk associated with not asking individuals about incidents of abuse (Becker-Blease and Freyd, 2006).

The feminist and queer theoretical paradigms were central to the development and conceptualization of the study, and to the way in which the research was conducted, providing a source of guidance for the numerous challenges faced. In thinking about how to conduct research with this minority and vulnerable population in an ethical manner, the use of a feminist and queer paradigm, provided an anchor for how to report the results of the research without contributing to and fuelling stereotypes of community dysfunction. Although not seeking to advocate for the particular paradigms adopted, having a defensible theoretical anchor is imperative to responding to challenges faced when conducting such research in a consistent and ethically uncompromising way.

From the outset the authors sought to be reflexive and responsive to the needs of the target research population. Cognizant of the multiplicity of identities, the term LGBTIQ was conceptualized as a marker and not as a defined category. Participation was open to anyone who identified with the LGBTIQ community without concern to the particular label that they adopted as an individual. On reflection, further thought should have been given to allowing participants to make an autonomous choice to participate while acting to minimize the preconceptions concerning the terms victimization, harassment, bullying and/or violence. In attempting to be transparent as to the topic of interest, the authors likely failed to capture a number of participants, who, although keen to participate, did not respond to the advertisements because they felt their experiences were insufficiently serious and thus excluded them from participation. The result is research that, although valid and reliable, does not holistically display the breadth of experiences of the community.

In operationalizing queer methods and empowering students to self-define their identity, the authors faced the challenge of how to present the research findings, utilize participant self-labelling and ensure the confidentiality of participants. A pragmatic and case-by-case approach appeared to be the most appropriate response. The paradigms within which the research was conducted require the identification of factors including race and disability that may lead to victimization and harm. The challenge was thus: to ignore them would be to conduct poor research; but to detail them would potentially lead to identification and participant unsafety. The outcome was an attempt to balance safety and confidentiality in reporting categories and intersections of victimization appropriately, ensuring that narratives were not erased or homogenized.

Qualitative researchers, and in particular those in critical paradigms, should think about the limitations created by failing to question the institutional review processes rooted in positivistic research paradigms. Research undertaken with human participants should try to maintain sufficient flexibility to respond appropriately to issues that arise when utilizing methods that place value in engagement with subjective data. In addition, through designing projects to be deemed ‘ethical’, researchers can find themselves responding more to the real (or perceived) threat of external oversight, with the effect of creating a research project that places unintended obstacles in the way of the research that we want to undertake. As Jennings argues, ‘ethics review can be helpful to researchers … they need to own it and use it as a tool for improving their practice … not treat it as a compliance hurdle to be overcome’ (Jennings, 2012: 94). Here, in unquestioningly providing participants with details of how to seek help, this resulted in increased risk to a participant. The challenge was thus how to meet institutional ethical requirements, while not creating a risk to participants’ safety.

An overriding concern was to not generate further abuse when conducting research with participants. In recording disclosures of violence and in hearing of incidents, perpetrated by members of a community to which the authors belong, there is a dilemma in how to best respond to the incidents reported and to prevent future acts. Though there is no single correct response to resolve the conflicts that arose, the theoretical paradigm was utilized to guide the responses and through engagement in the research process empower the participants, in cases where they may have been previously disempowered.

The authors were constantly reminded of the need to question and challenge orthodox approaches when working with minority and vulnerable populations. Many participants indicated their desire to be part of the project to facilitate change, even though they may not necessarily have seen themselves in an activist role. Importantly, each interview provided a reminder that the research participants are people; their thoughts and feelings became tangible; they also saw themselves as stakeholders for change and used the opportunity to participate in tandem with researchers, to impact society and see their hopes for change realized.