Sexual Violence Against Men in Conflict and Post-Conflict Settings: A Qualitative Research Methodology

Abstract

Sexual violence against men in conflict and post-conflict settings is under-researched. Men’s reluctance to talk about their experiences in conflict and post-conflict settings has contributed to the subject being a difficult area of inquiry. This article describes the research design and the strategies employed by the first author, who conducted qualitative research with South Sudanese male refugees who were survivors of sexual violence and have resettled in two communities in Uganda since the onset of the 2013 South Sudan conflict. The qualitative research is a part of a mixed-method program of research for the first author’s doctoral research that focused on sexual violence perpetrated against men in conflict and post-conflict settings. Based on the learnings during the fieldwork in this complex research setting and by drawing on best practices in qualitative research, this article proposes guidelines that can assist researchers who conduct qualitative research with vulnerable populations (across multiple disciplines) on sensitive topics such as sexual violence. The guidelines include five key steps: spending time in the community before participant recruitment and data collection, fostering a trust relationship with stakeholders, using appropriate gatekeepers, making participants feel at ease throughout the research, and using the snowballing sampling technique. The key steps are interdependent and can be adapted to suit the research context. These guidelines can be useful across multiple disciplines and subject areas.

Introduction and Background

Throughout ancient and modern history, sexual violence has been frequently reported in armed conflicts, to the extent that its occurrence is often considered a disturbing, yet inevitable outcome of armed conflicts (Leatherman, 2007; Meger, 2010; Oosterhoff et al., 2004). For example, about 500,000 women were reported to have been raped during the Rwandan genocide (1994), and 64,000 women were raped during the Sierra Leone civil war (1991–2002; Leatherman, 2011). Women are not the sole victims of sexual violence in armed conflicts as sexual violence against men have been documented in more than 25 different armed conflicts globally, and the subject has gained more attention from scholars in the last decade (Apperley, 2015; Keygnaert et al., 2014; Sivakumaran, 2007; Solangon & Patel, 2012). Sexual violence adversely affects all aspects of the male survivors’ health: physical, mental, social, and sexual health. The physical health consequences of sexual violence against men include anal fissures and fistulae, rectal bleeding, fecal and urine incontinence, groin pain, genital scarring, castration and swollen and painful testicles, and sexually transmitted infections (STIs) (Christian et al., 2011; Chynoweth et al., 2020a; Loncar et al., 2010; Norredam et al., 2005). The mental health consequences include suicidal thoughts and attempts, depression, post-traumatic stress disorder, auditory hallucinations, memory loss, and sleeping disorders (Christian et al., 2011; Chynoweth et al., 2017; Johnson et al., 2008, 2010). Male survivors of sexual violence continue to suffer the health consequences of the violence when they arrive in refugee resettlement communities and possibly long after the conflict is over.

According to the United Nations Office for the Coordination of Humanitarian Affairs (UNOCHA, 2015), humanitarian responses are initiated during emergencies like armed conflicts, to save lives, alleviate suffering, and support survivors. Although the Inter-Agency Standing Committee, the body that co-ordinates humanitarian responses in complex emergencies, makes considerations for male survivors in discussions on how to support sexual violence survivors, the majority of existing support programs for survivors of sexual violence are targeted toward women and children (Christian et al., 2011; Sivakumaran, 2007; Spangaro et al., 2013; Tol et al., 2013). Scholars have called for more studies to understand sexual violence against men in conflict, to provide evidence that can inform the development of services to support male survivors (Broban et al., 2020; Chynoweth et al., 2017). The methodological aspects of existing literature on sexual violence against men in conflict need to be improved. Some scholars on sexual violence against men depended predominantly on information provided by sources other than the male survivors of sexual violence in conflict themselves. An example is the reliance on information that is obtained from male sexual violence survivors in peacetimes and female sexual violence survivors in conflict, to describe the experience of men in conflict (Kinyanda et al., 2010; Sivakumaran, 2007). As Ó’MÓChain (2016) points out, the forms and motivations for male and female sexual violence victimization may be different, as is sexual violence perpetrated during peacetime compared with conflict situations. Similarly, reliance on information provided by humanitarian agencies’ staff (Manivannan, 2014; Sivakumaran, 2007, 2010) can be inaccurate as the information provided by them, although valuable, is prone to organizational bias and may be incomplete. Scholars have cited difficulties recruiting male sexual violence survivors, which are due to the associated shame, stigma, and socially construed expectation of masculinity, as the reasons researchers often rely on other sources of information (Christian et al., 2011; Clark, 2017a; Sivakumaran, 2007). The World Health Organization (WHO, 2007) suggests that sexual violence survivors should not be interviewed directly, except when there is no alternative. We believe that allowing male survivors share their lived experience with researchers is crucial to understanding the problem and using this evidence to develop appropriate interventions and strategies to support male survivors of sexual violence. Some scholars that have included men as participants in their studies relied on focus group discussions (FGDs) to obtain information from participants (Chynoweth, 2017; Chynoweth et al., 2020a, 2020b; Schulz, 2021). One of the limitations of FGDs is that participants may be unwilling or uncomfortable to share their detailed experience in a group, perhaps with strangers (Liamputtong, 2007), especially for a sensitive subject like sexual violence.

A qualitative study, which formed part of an exploratory sequential mixed-methods research project, was conducted to explore and understand the forms and the health implications of sexual violence perpetrated against men in conflict and post-conflict settings. The qualitative study included male sexual violence survivors as the primary participants. The male sexual violence survivors who participated in the study were not receiving support by any agency to deal with their experience of sexual violence, at the time of the study. Individualized, in-depth semi-structured interviews with male sexual violence survivors provided an opportunity to obtain rich qualitative data on this sensitive subject from the survivors themselves. In this article, we propose five “best practices” (referred to as five steps) for recruitment and data collection in qualitative studies with a vulnerable population, that explores a sensitive topic. The five steps include the following: spending time in the community prior to participant recruitment and data collection, fostering a trust relationship with stakeholders, using appropriate gatekeepers, making participants feel at ease throughout the research process, and using the snowballing sampling technique. These steps formed the conceptual framework for our research activities in the field (described in the second section). We also include a discussion on how researchers can employ this approach when conducting studies that focus on other sensitive subjects with participants who would rather remain anonymous.

Five Key Steps for Conducting Qualitative Research With Vulnerable Populations on Sensitive Subjects

Step 1: Spend Time in the Community Prior to Recruitment

During a study with an indigenous population in the United States, Burnette et al. (2014) developed a toolkit for conducting culturally sensitive research and one of the items in the toolkit is the need to spend time in the field to understand the people being studied. This is a common practice in qualitative research as it provides an opportunity to learn about the culture, values, beliefs, languages, norms, and interests of research participants; a critical step in conducting culturally sensitive studies in public health (Burnette et al., 2014; Flick, 2018). Cultural sensitivity in research is a key ethical issue, especially when conducting research with vulnerable populations. Understanding the unique culture of the population involved in a study is critical to the success of the research (Berg, 1999; Gostin, 1995). Cultural sensitivity in research increases the credibility and suitability of the research findings in proffering solutions to public health problems because the solutions developed from such studies consider population-specific determinants (Burnette et al., 2014).

Step 2: Develop a Trust Relationship With Community Leaders and Stakeholders

Building a trust relationship early is crucial in research with a vulnerable population because historically, the level of trust between some vulnerable populations and researchers is minimal; study populations often repute the word “research” as “dirty,” because there is a history of perceived or actual exploitation or harm by researchers to participants (Balestrery, 2010; Burnette et al., 2014; Deloria, 1991). The distrust that vulnerable communities have for researchers could explain an experience that Clark (2017b) had during participant recruitment in a study that explored the use of rape in the Bosnian war. Clark had solicited the head of the Croatian Camp Survivor Association to help researchers identify survivors of sexual violence, then Clark was informed by the head of the association that the survivors would be disinclined to share their experiences with researchers. However, Clark discovered that this claim was false as the survivors who were eventually recruited into the study, using a snowball sampling strategy, were willing to share their experiences. A possible explanation was, the head of the association told Clark that survivors would be unwilling to share their experiences because the leader felt an obligation to protect sexual violence survivors from researchers with whom they had no prior relationship. Community leaders’ tendency to try to protect the vulnerable populations they serve has been documented in the literature (Cramer et al., 2016; Liamputtong, 2007). A key strategy to overcome such an obstacle is to develop a trust relationship with the stakeholders and community leaders before commencing recruitment and data collection.

Step 3: Use Appropriate Gatekeepers

Gatekeepers are persons who occupy strategic positions within a community and can help researchers gain access to potential study participants (Liamputtong, 2007; Weissinger, 2020). To be considered an “appropriate” gatekeeper, the person must be able to help researchers access participants that possess rich information on the subject under consideration. The choice of appropriate gatekeepers can help researchers to manage their time and limited resources as they generally have easier access to the participants. A typical example of an appropriate gatekeeper is the Snaga Zene, a nongovernmental organization (NGO) that helped Clark (2017a) to recruit female participants into the study on rape in the Bosnian war. Snaga Zene is an NGO that supports female survivors of sexual violence, and they were able to contact some of their clients to participate in the study. The study had a disproportionately large number of female participants compared with males; one reason for this was the difficulty Clark experienced accessing male sexual violence survivors (Clark 2017c). The situation could have been different if there was a similar NGO in the study site that supported male sexual violence survivors (Clark, 2017a). Schulz (2021) employed the assistance of the Refugee Law Project, an NGO that supports refugees in Uganda, to recruit male sexual violence survivors into their FGDs. Recruiting participants through NGOs may not always be possible considering that there are not many such organizations that support male refugees specifically, and that the male sexual violence survivors themselves may be reluctant to approach the existing organizations (Chynoweth et al., 2017; Spangaro et al., 2013; Tol et al., 2013). Hence, it is important to consider other possible gatekeepers in such studies.

Step 4: Making Participants Feel at Ease Throughout the Research Process

Research participants can be distrusting of researchers for the same reasons as their community leader, discussed in Step 2. Therefore, developing rapport with participants should be one of the first steps undertaken by researchers when they enter the field (Guillemin et al., 2018; Pitts & Miller-Dau, 2007) and this needs to be maintained throughout the research process. Trust building is central to making participants feel at ease, especially in studies on sensitive subjects. Scholars have noted that gatekeepers play a vital role in the establishment of trust with research participants and making them comfortable to share their experience with the researcher (Belliveau, 2018; Lutfun, 2020). Another step that helps to build trust and make participants feel at ease is to honestly discuss the details of the research with participants, including the risk (and strategies to mitigate such), benefits, plans for privacy and confidentiality, ethical approvals granted, and expected outputs from the research (WHO, 2007). Hansen (2006) suggested that allowing participants to choose the location of where they engage in data collection, and to commence interviews with a general discussion, would allow participants to feel comfortable.

Step 5: Use Snowballing Sampling Technique

Different forms of purposive sampling methods are favored in qualitative studies, depending on the research question and objectives, but when the participants are considered difficult-to-reach, snowball sampling is the preferred purposive sampling technique (Hansen, 2006; Liamputtong, 2007). In snowball sampling, a participant nominates other people they know who have similar experiences, as potential participants. This sampling technique can be successful only if participants are aware of other sexual violence survivors.

Research Design

Study Location and Participants

The qualitative study was conducted between January and April 2019, with South Sudanese males who resettled in Imvepi and Rhino resettlement communities in Arua district, Uganda, since the onset of the 2013 South Sudan conflict. Thirty-two in-depth semi-structured interviews were conducted and 26 were with South Sudanese men who had experienced sexual violence in conflict and/or in post-conflict resettlement communities. Six interviews were conducted with humanitarian aid workers employed with United Nations High Commissioner for Refugees (UNHCR) and NGOs that provide support for survivors of sexual violence living in the resettlement communities.

Recruitment of Study Participants

We employed the snowball sampling technique, a purposive sampling strategy, to recruit the male survivors of sexual violence. Community and youth leaders in the two resettlement communities acted as gatekeepers and they assisted us to identify and recruit the first seven participants to the study (see “Research Strategies” below). To avoid anchoring, we conducted the interviews in multiple (seven) villages within the two resettlement communities.

Data Collection and Management

The in-depth semi-structured interviews conducted with the 26 male survivors and six humanitarian workers used a question guide, a different guide with each cohort, and all the interviews were audio-recorded, with the verbal consent of the participants. Nineteen interviews were conducted in English and a male research assistant (RA) served as an interpreter during the seven interviews when a participant preferred to speak Arabic. The RA had been living in another refugee community in Uganda for 5 years and had good knowledge of the South Sudanese culture and languages spoken, and the terrain of the resettlement communities. All interview transcripts were translated into English and safely stored on password-protected computers and the hard drive was accessible only to the research team.

Research Strategies

Selection of Study Population

The choice of the study population was deliberate as it was important to have access to a sizable number of information-rich male sexual violence survivors in an environment that was safe for the research team to undertake data collection. At the time of the research, the crisis in South Sudan was the largest refugee crisis in Africa and the third largest in the world, after Syria and Afghanistan. We opted for South Sudan refugees (rather than refugees from Syria and Afghanistan) because the researcher is a male of African origin, which provided him with relatively easy access to the African population. Although it would have been possible for someone of non-African origin to effectively conduct this program of research in Africa and vice versa, the choice of location for this study, where the researcher felt connected to the host community, certainly made the participants feel more comfortable. For example, one of the participants told the student researcher that “I feel this comfortable talking to you because you are one of us (pointing to his skin).” During the fieldwork, it was easy to relate to the South Sudanese cultural values as they are similar to those of the student researcher.

We included humanitarian aid workers as key informants, as they had been providing support for survivors of sexual violence, and could provide an insight into the subject from a service provider’s perspective (Liamputtong, 2007). The inclusion of the aid workers in the study was important because it revealed different perspectives than that of the survivors regarding sexual violence against men in conflict and post-conflict settings. In addition, it confirmed our initial belief that studies that rely solely on information provided by service providers, such as humanitarian aid workers, are unlikely to gain a complete understanding of male survivors’ lived experiences of sexual violence.

Selection of Study Sites

The selection of the study site was also a deliberate strategy as the aim was to achieve a high yield of information-rich participants. According to a UNHCR (2017) report, an estimated 2.43 million South Sudanese have been displaced into neighboring countries like the Republic of Uganda, Kenya, Sudan, Ethiopia, and the Democratic Republic of Congo, since the onset of the 2013 South Sudan conflict. The key factor that informed the choice of Uganda as the study site, was that the Republic of Uganda borders South Sudan to the south and accommodates the largest percentage (about 42.6%) of the South Sudanese refugees (UNHCR, 2017). Uganda ranks fifth globally and first in Africa among countries receiving refugees and has refugee-friendly policies in place to ensure the refugees settle into host communities with more ease, by providing them with land and employment opportunities (Amnesty International, 2020; Glass & Doocy, 2013). During a scoping visit to Uganda in May 2018, we learned that the country encourages research that has the potential to improve the country’s refugee protection policies and programs, which further supported the choice of Uganda as the study site.

In Uganda, we opted for the West Nile district of Arua, for two reasons. First, during the scoping visit to Uganda, the Directorate of Refugee and Disaster Preparedness in the Office of the Prime Minister informed the research team that they would only be granted approval to conduct the research in a single district. Although the initial plan was to go to the largest resettlement community in the country (Bidibidi in Yumbe district), we realized that two of the five largest resettlement communities, Rhino and Imvepi, were in the same district of Arua. We opted for Arua district and resettlement communities because having two communities in one district increased the chance of recruiting enough male participants that had experienced sexual violence in conflicts and/or post-conflict settings. Second, the safety of the researchers, which is an important ethical issue in this type of research (Liamputtong, 2007), influenced the choice of Arua district. Considering that there was an ongoing war in South Sudan at the time of the research, we opted for a district that had a high South Sudanese population that was farthest from the border of Uganda and South Sudan.

Recruitment of Participants

The research team moved into Arua district about 4 weeks prior to commencing participant recruitment, to lay the foundations for the fieldwork in the resettlement communities. This is the first of the five steps described above and the activities during this period contributed to the success of the research. We were unable to make direct contact with participants at this stage, as we were awaiting ethics approval from the Ugandan National Council for Science and Technology (UNCST), so we focused on engaging the community and youth leaders in discussions, as a means of developing a trust relationship with them and their community. The RA, who was very familiar with the resettlement communities, was instrumental in arranging meetings with the elected leaders in the community. This period provided an opportunity to learn about the South Sudanese men’s culture, beliefs, interests, and traditions, so that we could carry out culturally sensitive research with this vulnerable population (Wilson & Neville, 2009).

Initially, the community and youth leaders were reluctant to engage with us, but they became more receptive once we explained the rationale, benefits, and possible adverse effects (and our plan to mitigate such) of our research. The leaders became even more comfortable and interested in the research when we demonstrated our interest in learning about their culture and beliefs (see Step 2). Then, they enthusiastically told us about their country, tribes and taught us some words in Arabic, which proved useful during data collection.

The importance of building a trust relationship with participants was later highlighted in two separate events. The first occurred when the secretary of a village in Rhino community initially disallowed us from interviewing a man in his village because he was not consulted prior to the day of the planned interview. However, after speaking with other village leaders who we had met earlier, the secretary allowed us to conduct the interview; this was the first time we met the secretary of the village as he was not in the village during our previous visit. The second relates to a male survivor of sexual violence who, during the post-interview conversation, said how much he appreciated the research team for “coming down to their level” and relating with them like fellow human beings. He noted that he had not experienced this interaction with some previous researchers who he claimed had treated them like samples in an experiment.

We commenced the recruitment process when we were confident that the community leaders trusted us. During the recruitment process, we needed the help of gatekeepers, who would help us gain access to the participants, as explained in the third step above. The initial research plan was for the community/youth leaders and humanitarian aid workers to act as gatekeepers because we believed that they understood the challenges faced by refugees and know which men in the resettlement communities had experienced any form of sexual violence. The community and youth leaders were enthusiastic and active, and they worked with us to recruit male survivors of sexual violence. Conversely, the humanitarian aid workers were unable to nominate any male survivors to participate in the study, despite stating they were willing to do so. The major reason the humanitarian aid workers provided for not being able to nominate any potential participants was because they had only a few male sexual violence survivors attending their services.

The community leaders and humanitarian aid workers were asked to distribute a flyer containing information about the research and the contact details of the RA (a dedicated phone number for research purposes), to men they considered to have sufficient information on the issue. The men who were interested in taking part in the research contacted the RA directly, rather than the student researcher. This strategy offered a two-way protection for the participants’ privacy and confidentiality. First, the identity of survivors who were contacted by the community leaders but refused to participate were never known to the researchers. Second, the identities of participants were protected from the gatekeepers because they contacted the RA directly. When the potential participants contacted the RA, they were given additional information about the study, using the plain language statement, and they were allowed one week to consider their participation. If they phoned the RA to inform him that they were willing to participate, then we obtained their verbal informed consent and proceeded with data collection. The verbal consent was audio-recorded, and the participant’s pseudonym was noted the codebook developed for the research. Both the codebook and the recording were stored in a password-protected device that is accessible only to the student researcher. Participants acknowledged that they became more comfortable with the researchers after we provided the details of the research, allowed them to decide if they want to participate and explained how privacy and confidentiality will be protected (Step 4).

At the end of each interview, we asked the participants if they were aware of other male sexual violence survivors, and if so, we asked them to give them a flier; we repeated the process stated above, in line with the snowballing technique described in the last of the five steps above. An important finding that relates to the recruitment strategy was that male sexual violence survivors discuss their experience with one another as well as with the community leaders and religious leaders. This was helpful in the snowballing process as most of the participants knew someone else that they could invite to the study, with the flier. At the completion of the interview, we referred all participants to an NGO that supports sexual survivors within the resettlement communities, for counseling. This strategy mitigated any possible re-traumatization from sharing their experience with the researchers.

Data Collection

The need to make the participants at ease, as described in Step 4 above, guided all activities during the data collection phase of the research. The interviews were conducted in locations that were considered comfortable and safe for both the participants and the research team, which in most cases, were outside the participants’ homes. Initially, we anticipated that the participants may be nervous and uncomfortable to talk to us about their experiences (Hansen, 2006), so we took steps to make participants feel more at ease. One way we achieved this was to warmly greet them and ask, “how are you doing” in Arabic. Greeting the participants in their language was very effective as they often smiled and asked if we were from South Sudan, or they asked how we knew their language. Before we commenced an interview, we talked briefly about something they considered to be interesting (this was an icebreaker). During our immersion in the community, we learnt that most South Sudanese men love soccer, so our icebreaker subject was often a discussion on the English Premier League or the Spanish LaLiga. Fortunately, the student researcher is a soccer enthusiast, which helped the researchers engage in a conversation around soccer. During one interview, a participant wore his soccer club jersey and co-incidentally, the researcher wore his jersey the same day. The “connection” that developed throughout the interview between the researchers and the male participant was tangible. During all interviews, we demonstrated genuine empathy with the participants as they talked about their experiences of sexual violence.

During interviews, we did not launch into questions about participants’ sexual violence experiences. Rather, after collecting the socio-demographic information, we asked the men about their well-being and their general experiences during the war in South Sudan, during flight and since arrival in the resettlement community. Apart from easing into conversations on sexual violence, the slow immersion into the interview process proved to be useful in other ways. One example was, the men talked about some of their personal experiences that aligned with the definition of sexual violence, yet the men had not considered them as forms of sexual violence. For instance, while discussing their general experiences during flight, several participants spoke of being stripped naked or beaten on their genitals, without considering such acts as a form of sexual violence.

The in-depth interview guide for the male survivors consisted of four main questions with associated prompts; however, we were flexible with the order of the questions and allowed for a natural flow in the discussion. During the interviews, we explored other issues raised by the men that we considered were relevant to the research topic. We conducted follow-up interviews with 10 participants to seek clarification or to explore an issue that had emerged during the initial interview, in more depth.

During the interviews, we experienced some complications with two participants, which we had anticipated. We were aware that discussing sexual violence could be an unpleasant experience for some participants and male survivors may feel emotional pain and become distressed when talking about their experiences. The two men became visibly agitated and cried during the interview as they talked about their experience of rape during flight. When we noticed their agitation, we stopped asking questions, reassured the men of our intentions and their confidentiality, and reminded them of their rights to withdraw from the interviews without any consequence. We said that they could terminate the interview if they wished, or arrange a more suitable time to recommence their interview. In both cases, the men eventually calmed down and they agreed to continue with the interview. At the end of the interviews, both men, as with the majority of the participants, expressed their gratitude to the research team for conducting such a study, noting that participating in the research was therapeutic for them as they felt some relief after sharing their experiences. The first indication that we had reached data saturation, the point where no new themes were emerging, occurred after completing interview 21. However, we conducted another five interviews with male survivors to avoid turning back men who had already booked in for an interview; no new themes emerged from these five interviews.

Transcription of the recordings was done within 48 hr of the interviews taking place. Participants were invited to undertake member checking about a week after their interview, and all participated, except one man who had to travel to South Sudan soon after he participated in the interview. Each participant was given a paper version of the transcript of their interview to read. We asked the participants to confirm that their transcript conveyed the exact experiences that they shared with the researchers. They were invited to add or remove content as they considered appropriate.

Ethical Considerations

We obtained ethics approval from the Deakin University Human Research Ethics Committee (DUHREC) (2018-223), the Research Ethics Committee in the School of Social Sciences, Makerere University (MAKSS REC 10.18.227); the Directorate for Disaster Preparedness and Refugees in the Office of the Prime Minister (OPM) in Uganda (OPM/R/41/1) and the Ugandan National Council for Science and Technology (UNCST) (SS 4829).

We considered several ethical issues, the details of which are beyond the scope of this article but will be mentioned briefly here. First, we conducted the research, in the first place, because we believe that its benefit for the survivors and their family outweighed the risks associated with the research process. We ensured strict privacy and confidentiality for all study participants and obtained verbal informed consent before commencing data collection, as discussed above. We anticipated that when participants shared their experience, they may be re-traumatized, so we made arrangement for free of charge counseling with one of the organizations in the resettlement and referred participants to this service at the completion of the interview. As required by the WHO (2007) guidelines, we would share the findings of the study with the research participants. On the advice of the Research Ethics Committee in Uganda, participants were compensated for their time with 10,000 Ugandan Shillings.

Discussion

Spending time in the field prior to recruitment may seem cumbersome and a waste of time and resources, as it may appear that the researchers are not doing actual “research work” during this period. We recommend that researchers who are interested in conducting research with other vulnerable or “difficult-to-reach” populations, like refugee men, should consider this step as an integral part of the research design, because it has the potential to facilitate other research activities. During this period the researcher can build trust with the community leaders and other stakeholders, as was the case in our study. The way this was achieved in our study has been described in the research strategy. In addition to contributing to the conduct of ethical and culturally sensitive research, the immersion period provides an additional opportunity for the researchers to undertake participant observation which could yield additional rich information. If the study population is not in a well-defined geographical area like a refugee resettlement community, the process of immersion could be more difficult. In this instance, researchers should consider identifying associations or individuals that represent the population, like Clark (2017b) did, and invest time to build a trust relationship and to know more about the participants before commencing recruitment. We do not recommend a defined amount of immersion time, but suggest researchers only commence recruitment when they are convinced that a trust relationship has been developed. In our study, we knew this had occurred when the community leaders were collaborative and enthusiastic about the research.

The choice of appropriate gatekeepers is critical to the success of research with vulnerable populations about a sensitive subject like sexual violence. Whereas researchers often approach NGOs and humanitarian agencies to help them recruit participants (Chynoweth, 2017; Clark, 2017a; Schulz, 2021), the absence of NGOs that support men and the reluctance of survivors to approach the agencies for support are limitations to such a recruitment approach in some settings. In our study, we encountered a similar situation where the NGOs who were initially considered to be primary gatekeepers were unable to nominate a single participant throughout the study period. Although they were enthusiastic about helping us, the NGOs were unable to nominate participants because their clients were mostly females. During the data collection phase of this study, we confirmed our earlier suspicion that survivors talk to one another and to their community leaders about their experiences of sexual violence, which justified our decision to use the community leaders as gatekeepers. When determining the most appropriate gatekeepers in a study, an important question to consider early in the study design phase or during the scoping visit (if this occurs) is, who are the survivors most likely to trust in their community? People are more likely to share their experiences with people they trust, especially on sensitive issues (Mittendorf et al., 2019) and we confirmed this to be true in our study. With the benefit of hindsight, we now know that religious leaders in our study sites would have been appropriate gatekeepers because as noted earlier, we found that sexual violence survivors discuss their experience with their religious leaders. Although male sexual violence survivors may be reluctant to admit their experience to staff of humanitarian agencies because they consider them as “outsiders,” they often share their experience with one another and with the people in their community that they trust, like their community and religious leaders. This is not to say that humanitarian agencies and NGOs’ staff should not be used as gatekeepers, but it is important to consider other possible gatekeepers in addition to those commonly employed for participant recruitment.

The establishment of trust relationships, which commenced during the 4 weeks of immersion in the community was the single most significant factor that contributed to the success of this study. We consider that investing time and effort into developing genuine trust relationships with community leaders and stakeholders is critical for obtaining a high yield of information-rich data, for both ethical and strategic reasons. The ethical consideration was, we could not contact sexual violence survivors without the consent and support of community leaders and other stakeholders (WHO, 2007). The strategic reason was, we anticipated that the trust relationship we developed with the leaders would foster the development of a trust relationship with the participants. A study with an indigenous population in Australia reported that the more research participants know about and trust a researchers’ institution (university), the more likely they are to talk to researchers (Guillemin et al., 2018). If participants can trust researchers because of their affiliation to a university, an institution where the participant may not know anyone, they are more likely to be at ease with researchers who approach them based on the recommendation of the trusted, closer community association. This was the case with our study, and we recommend that developing a relationship with community leaders should precede contact and relationship with participants.

It takes courage for survivors to discuss their experience of sexual violence. Although it is an ethical requirement that participants volunteer information to the extent that they are comfortable (WHO, 2007), making participants feel at ease is essential to ensuring that participants tell their story voluntarily and in considerable detail. As noted above, participants do not want to feel like subjects in an experiment, but they prefer researchers that make them comfortable. There are steps that researchers can take to make participants comfortable. In our study, the identity of the student researcher as an African male played a role in making the participants comfortable. Although this may not always be the case, researchers should consider having someone on the team that shares some characteristics with research participants. This sense of being “someone like us” will make male sexual violence survivors comfortable, especially when as mentioned earlier, they are aware that the research team is known to the community and religious leaders that they trust. The use of some common words in the participants’ language, and relaxed conversations about participant’s interests, will facilitate a more relaxed atmosphere during the interviews (Ellard-Gray et al., 2015; Shedlin et al., 2011), as occurred in this study. Researchers can learn about participants’ language and interests during the immersion period. Furthermore, a detailed, honest description of plans to ensure privacy and confidentiality for participants and allowing the men to choose a location where they feel safe, for the interviews, will make participants feel more comfortable. It is noteworthy that the level of trust and comfort that participants have with the researcher would determine whether they are willing to nominate other participants in the snowballing process.

Conclusion

The goal of humanitarian responses is to support survivors of humanitarian emergencies such as conflict and disasters. Some of the survivors have suffered horrifying experiences including, but not limited to, sexual violence. The most appropriate policies and programs to support survivors of sexual violence need to rely on quality empirical evidence. We believe that the “owners of an experience” are best placed to provide the most accurate data/evidence on a subject that relates to their experience.

The international literature notes that men are reluctant to discuss their experience of sexual violence in conflict and post-conflict settings because of their perception of masculinity and the shame associated with emasculation. In the past, most of the discussions on sexual violence perpetrated against men in conflict and post-conflict settings have been based on information provided by sources other than the male sexual violence survivors themselves. However, we believe that it is possible to obtain high-quality information from men about their experience of sexual violence.

In this article, we draw on five best practices (referred to as five steps) in qualitative research to highlight strategies that qualitative researchers can employ to obtain quality data on sensitive topics, such as sexual violence with difficult-to-reach populations like male survivors in humanitarian settings. We emphasize the importance of spending time in the community prior to participant recruitment and data collection; fostering a trust relationship with stakeholders; using multiple appropriate gatekeepers; making participants feel at ease throughout the research; and using the snowballing sampling technique. The “Research Strategies” section of this article can provide a comprehensive guide to researchers interested in similar research in the humanitarian and other settings. These steps discussed are interdependent, and researchers need to adapt each step to suit their local context.