Methodological Challenges in Conducting Sexual and Reproductive Health Research Among Young Males in Bangladesh: Reflections From a Nationwide Mixed Methods Study

Abstract

Applied mixed methods research for sexual and reproductive health (SRH) issues involves multiple researchers and presents numerous responsibilities as well as unforeseen challenges. These challenges together with the often ignored and under-researched area of male sexual and reproductive health and rights (SRHR) in Bangladesh require researchers to navigate cultural stigmas associated with SRH and hostile political situations across the country. We conducted a nationwide mixed methods research study to explore the SRHR of male youth. This study is the first of its kind in the context of Bangladesh. The aim of this paper is to describe the challenges experienced and mitigation strategies undertaken to conduct research on the SRHR of young males in Bangladesh. A group of experts, practitioners, service delivery staff and researchers were consulted during the tool development stage. However, challenges in obtaining permission from community members, explanation of SRH-specific topics, and establishing respondents’ trust persisted throughout the study. Mitigation strategies included spending time outside of scheduled interview hours and making repeated visits to respondents to establish trust. These strategies allowed for honest discussions over time and served to assure our study respondents that the confidentiality of their data would be protected. In addition to sourcing letters from official authorities, we had conversations with community influential people such as religious leaders and political representatives about the nature of our study and explained to them the implications of the study for the future of young men’s health. Moreover, we highlight the political issues of the time and shed light on the ad hoc measures taken to overcome the challenges we faced. These reflections can provide valuable insights for researchers conducting SRHR-related studies in similar social and political settings.

Introduction

In the last decade, there has been a growing interest in the field of sexual and reproductive health and rights (SRHR). SRHR has been prioritized in the Sustainable Development Goals (SDGs 3, 4, and 5) which are committed to “leaving no one behind” (United Nations, 2015, p. 5). Within the broad scope of SRHR, the sexual and reproductive health (SRH) experiences and needs of male youth (hereafter referred to as young males) aged 15–24 years have traditionally been ignored (Marcell et al., 2011).

Less than 41% of adolescents use contraception in most countries, with male condom use at only 5.5% (The World Bank, 2015; UN Department of Economic & Social Affairs, 2019). In Bangladesh, Ahmed et al. (2009) found that male university students did not receive proper information on sexual and reproductive health; and instead, they were discouraged and criticized when seeking such information. About 48% of the respondents did not have a clear understanding of puberty, and they discussed puberty issues with their peers (84%), which made them vulnerable to inadequate and incorrect information. The same study also revealed shyness and confusion on matters related to sex and sexuality, which has implications on young men’s emotional well-being, as well as that of their partners. Moreover, a study using Bangladesh Demographic and Health Survey data by (Bishwajit et al., 2017) showed that only 40% of men were involved in active family planning (FP) and had low awareness and knowledge across SRH topics. This is a trend in South Asia, where knowledge of contraception is highly reported, but details of SRH processes is limited. A study in two districts of Madhya Pradesh India reported high knowledge of condoms but had limited SRH knowledge (Char et al., 2011) . Another study on SRH in India, Nepal, and Bangladesh by the International Planned Parenthood Foundation (IPPF), found that men reported traditional beliefs related to causal factors of SRH problems, where in Focus Group Discussions (FGDs), Bangladeshi men perceived SRH issues as beyond the control of humans; traditional beliefs include: ‘having sex after drinking alcohol results in sexual impotency and also produced weak child’ (IPPF, 2013). Furthermore, there was a negative perception of condoms as it reduces sexual pleasure, and the majority of men in Bangladesh reported that contraception was a ‘woman’s business’, and that men need not worry about it (IPPF, 2013). The South Asian example also indicates the importance of a cultural meaning system, whereby the concept of family in Asian societies takes precedence. A study on South Asian men living in the United Kingdom (UK) revealed that they seek help for issues such as sexual dysfunction and urinary tract infection from their families before seeking professional medical help (Idris et al., 2017). In this paper, the authors concluded that social and psychological factors mediate masculinity, and health-seeking behaviour and utilization of health services are highly context-dependent. As a result, worries and vulnerabilities in terms of SRH may be similar, but support systems and health seeking practices will vary across socio-economic contexts. Also, a review of client register books in two randomly selected Adolescent Friendly Health Corners (AFHCs) established by the Government of Bangladesh, revealed that among all adolescent clients, only 2.5% of service-seekers were boys (Ainul, Ehsan, et al., 2017). Therefore, it is found that attitude towards formal SRH information and service utilization among young males are very limited in the context of Bangladesh. (Bishwajit et al., 2017)

Adolescents and youth (aged 15–24 years) comprise around 20% of the total population in Bangladesh (total population of Bangladesh is about 165.57 million, estimated as of January 2019) and almost half of them are young males (BBS, 2019). Data of the Government of Bangladesh (GoB) from 2019 revealed that almost 38% of couples used oral pills as a form of family planning, whereas condom usage among couples is only around 7.5% (Islam et al., 2020). Azim et al. (2008) found that negative societal views of unmarried men buying condoms are seen to persist in Bangladesh. Other reasons for low contraceptive use cited in the same study include personal pleasure and trust, where one respondent stated that condoms interfere with a “pure” and “natural” sex act (Azim et al., 2008, p. 317). However, reasons for opting for contraception and subsequent health behaviours have not been explored. Contraception often overlaps with sexual violence and reproductive coercion, which are controversial areas of discussion in the context of Bangladesh. Positive associations between condom use and education have also been observed in Bangladesh (Haseen et al., 2012). Although the benefits of sex education are well-documented, SRHR content is not taught properly in schools in the country due to teachers’ reluctance to discuss such topics (Ainul, Bajracharya, & Reichenbach, 2017).

Despite risky SRH behaviour and lack of comprehensive knowledge regarding SRHR among young males in Bangladesh, there exists a dearth of systematic evidence on their SRH needs and service-seeking experiences. While increased research is needed to fill the knowledge gaps pertaining to the SRHR of young males in Bangladesh, the challenges anticipated by researchers owing to the local socio-cultural context may deter them from pursuing research in this field. Furthermore, very little is known about the practical challenges faced by researchers involved in male focused SRHR studies in Bangladesh. One of the reasons is that such studies are by and large non-existent in the context of Bangladesh. Familiarity with the challenges involved in carrying out SRHR-related studies in Bangladesh may alleviate the concerns and enhance the preparedness of researchers interested in this field. This article aims to describe the challenges experienced and the mitigation strategies undertaken by researchers involved in a nationwide mixed methods study on the SRHR of young males in Bangladesh.

Methodological Challenges of Conducting Research in Sex, Sexuality, and SRHR

Researchers have documented some of the challenges of conducting research in the field of sex, sexuality and SRHR. There is a culture of silence around discussion on SRHR in Bangladesh which largely stems from shame, stigma, and taboo associated with SRH topics (Camellia et al., 2020). Barthakur et al. (2017, p. 128) mentioned beliefs exist among Indian women that their expression of sexual dissatisfaction can be interpreted as them being “bad women” since it would connote that they are demanding sexual pleasure. Similarly, evidence shows that sociocultural norms on masculinity may prevent male respondents from disclosing experiences of sexual exploitation due to the fear of appearing effeminate (Family for Every Child, 2018). Findings from a study conducted in Cambodia suggest that some participants thought that the sexual abuse of boys was not as serious an issue as it is for girls (Hilton, 2008). The study also found that boys’ “externalised behaviour is rarely interpreted as a sign that they need help, and they suffer in comparison to girls” (Hilton, 2008, p. 145). These challenges highlight the deeply gendered cultural norms and beliefs associated with sex and sexuality in some societies.

The challenges prevalent in survey research on SRHR, are well-established. Low response rates and an overreliance on volunteer samples can negatively impact the representativeness, and therefore the generalization, of survey findings (Fenton et al., 2001). Moreover, participation bias and recall bias can lead to errors in measuring sexual behavioural risk estimates (Fenton et al., 2001). Participation bias occurs when certain characteristics are disproportionately found in respondents. For instance, in a cross-sectional survey of highly sexually active female sex workers in South Africa, Ramjee et al. (1999) reported underreporting of sexual behaviours. Importantly, there is no “gold standard” for self-reports, a common method of data collection on sexual behaviour, making it difficult for researchers to interpret the data meaningfully (Schroder et al., 2003, p. 117). The diary is considered to be the most accurate form of self-reporting because it eliminates the issue of recall bias (McLaws et al., 1990; Durant & Carey, 2000). Surveys that require respondents to read and write may exclude those without literacy and long questionnaires can sometimes result in poor quality data (Fenton et al., 2001).

In a mixed-methods study on the SRH knowledge and beliefs of young Sudanese refugees in Australia, Dean et al. (2012) faced the issue of interviewees arriving late to meetings. The authors ascribed it to the norms of polychronic cultures, wherein many activities are carried out simultaneously, as opposed to monochronic cultures in which one event occurs at a time. Therefore, extra time had to be reserved for data collection. In addition to the issues of cost, sample size and statistical power, a challenge that researchers collecting data on sexual orientation and gender identity deal with often is the reaction of participants upon being asked sensitive questions (Sell, 2017). Based on the researchers’ experiences in a mixed-methods study on the sexuality of Indian women survivors of breast cancer, Barthakur et al. (2017) raised numerous questions regarding the possibility of miscommunication between the researchers and interviewees. One of the questions the authors raised is related to the misalignment between the words “climax” and “orgasm” used in the questionnaire and participants’ understanding of these terms. Thus, researchers should be wary of making assumptions about the meanings of terms as respondents’ own understanding of these terms can vary from that of the researchers. Especially in studies related to sexuality and SRHR, researchers should be mindful of the fact that “views of which behaviours constitute having “had sex” also vary widely between individuals, and definitions of sexual activity can be influenced by cultural factors” (Ansara, 2015, p. 368).

The literature on methodological challenges of SRHR-related studies in Bangladesh is scant. However, a few challenges in this regard have recently been documented. Researchers have cited cultural sensitivity to SRH issues among Rohingya girls and women and difficulty in building trust with the Rohingya community in the refugee camps in Bangladesh as some major challenges (Ahmed et al., 2020). In another study which focused on the SRHR of persons with disabilities in Bangladesh, researchers faced difficulties in developing culturally appropriate tools and observed that some respondents took offence when they were posed questions about their SRH (Amin et al., 2020). The authors of the same study noted that in a few cases where they sought help from caregivers to elicit responses from participants who could not communicate themselves, the researchers had to refrain from asking questions about sensitive issues such as sexual desire and sexually transmitted disease/infection (STD/STIs) on ethical grounds, as these are respondents’ private thoughts and feelings.

The Study Context

Evidence on the SRH beliefs, practices, needs, service-seeking behaviour, and service utilization related barriers experienced by young males in Bangladesh is extremely limited. We conducted a nationwide study to generate evidence on the SRHR-related knowledge, perceptions, practices, and service-seeking behaviour of young males aged 15–24 in Bangladesh. We employed multiple methods of inquiry to uncover the social and cultural factors and gender ideals that influence the SRH behaviours of male youth. This study followed a sequential mixed-methods design and both quantitative and qualitative data were collected using validated, pretested tools. The Institutional Review Board of BRAC James P Grant School of Public Health at BRAC University evaluated and approved the study protocol which described the study design, tools, sampling and data collection process, and ethical issues (2018–043-IR).

Under the quantitative component of this study, we conducted a cross-sectional nationwide survey from the first week of May 2019 till the first week of August 2019. The survey questionnaire consisted of elements such as socio-demographic information, SRH knowledge, contraceptive use, SRH-seeking behaviour, as well as validated scales on gender equality, masculinity, and religious inclinations. The survey covered all 64 districts (289 rural, 44 urban and 37 semi-urban clusters) from eight administrative divisions of Bangladesh. Using a probability sampling technique, a total of 11,012 respondents (aged 15–24 years) were surveyed using a two-stage stratified random sampling technique by considering 24 strata (rural-city corporation-other urban times eight divisions) and 16 domains (rural-urban times eight divisions). Eight teams, each comprising of six data collectors and one supervisor, collected data using a structured questionnaire. To cover the hill tract areas, we recruited data collectors who could speak Chittagonian dialect of Bangla language so that they could communicate with the respondents from indigenous groups more effectively. The trained data collectors conducted face to face interviews which lasted for 50–60 min (each). These data collectors administered the questionnaire, asked for responses from respondents, and collected data using an electronic device (SurveyCTO software and Open Data Kit platform). On the same day, each supervisor randomly re-interviewed two respondents from each cluster using a shortened questionnaire to monitor the quality of data collected. This shortened questionnaire carried two questions from every section of the original questionnaire. The supervisor stated the purpose of this re-interview and convinced the respondent to manage time. Whenever inconsistencies were found, the respondent was re-interviewed at the end of the data collection period for the respective location. If any respondent refused to be re-interviewed, the supervisors were instructed to convince them. However, during the survey data collection, such incidents rarely occurred.

The purpose of ethnography, which tends to focus on smaller populations, is to gain in-depth insight into a particular area of interest, rather than achieve representativeness (Neuman, 2009). Thus, for the qualitative component of the study, we conducted a multi-site-based focused ethnography to obtain a deeper understanding of the ways in which the lives of young male respondents shape their SRH experiences, beliefs, and practices. Focused ethnography, also known as “mini-ethnographies” or “micro-ethnographies”, is an evolving methodology that is usually conducted with a sub-cultural group, deals with an issue within a specific context, and takes place within a short timeframe (Knoblauch, 2005; Rashid et al., 2019, p. 1–2; Wall, 2015). Traditional ethnography, on the other hand, focuses on cultural groups that are distinct from those of the researcher(s) and tends to be much more time-consuming (Rashid et al., 2019; Wall, 2015). A focused ethnographic approach would allow us to spend a sufficient amount of time with respondents to gain deeper understanding of specific aspects of their way of life and being (Higginbottom & Cruz, 2013) which, in turn, would lead to an in-depth understanding of their sources of SRH knowledge and influences on their SRH behaviour.

Following a focused ethnographic approach, qualitative data was collected over the course of 6 weeks, from October to November 2019. Eight trained researchers were assigned to four different locations across five districts in Bangladesh which were selected based on geographical and ecological variation to best capture heterogeneity of SRH knowledge and practices with respect to location. The following sites were selected: Roangchari of Bandarban district (mostly composed of hilly terrain and ethnic and tribal communities); Sholoshohor railway slum of Chattogram City Corporation (urban slum of a port city and industrial area); Shyamnagar of Satkhira district (rural area in a coastal region); and Jagannathpur of Sunamganj district (a rural area located in a haor, a wetland ecosystem). We expected that language barrier would be a problem in the South-Eastern part of the country, especially in the hill tract zones where indigenous community resided. In those areas, Chitagonian dialect (dialect of Chattagram) is mostly spoken and understood by all. Therefore, we recruited data collectors for these particular study sites, who could speak Chitagonian dialect of the Bangla language. So that they could communicate more effectively, especially with the respondents from indigenous groups in hill tract areas. We also recruited one researcher belonging to Marma community (an indigenous group living in Bandarban hill tracts area) for qualitative data collection.

A total of 40 case studies (10 cases in each site) and eight focus group discussions (FGDs) (two FGDs in each site) were conducted with male adolescents and youth whose ages were between 15 and 24 years. Altogether, 14 case study respondents were under 18 years. Eight in-depth interviews (IDIs) with parents (two IDIs in each site), 17 key informant interviews (KIIs) with community influential persons, gatekeepers, and formal and informal service providers (a minimum of four KIIs in each site), four community mapping, four historical timelines, and four occupational indices (one in each site) exercises were also carried out. The age of the selected key-informants varied from 30 years to 65 years. The key informants were formal government healthcare providers (4; three doctors and one community healthcare worker), traditional healthcare providers (4), NGO healthcare provider (1), informal healthcare providers (3), political leaders (3), and schoolteachers (2). In the case studies, respondents’ marital status, income, education, history of drug use, the prevalence of sexually transmitted infections or sexually transmitted diseases (STI/STD), number of sex partners, internet use, etc. were considered as the selection criteria. Potential respondents for case studies were selected from the existing quantitative dataset. In the event where respondents were unavailable or unwilling to participate from the existing survey dataset, the researchers selected new respondents based on the selection criteria from the same community.

Since very little research has been conducted on SRHR in Bangladesh where SRHR topics are perceived to be highly sensitive, evidence regarding the common challenges in SRHR research in the country’s context is severely limited. Researchers faced multiple challenges during the fieldwork for both the quantitative (survey) and qualitative (focused ethnography) components of this study. By describing the challenges and mitigation strategies adopted by the researchers in this study, we seek to aid SRHR researchers in Bangladesh and other culturally similar low and middle-income country (LMIC) contexts to develop methodological strategies to overcome potential challenges.

Researcher Positionality

All the researchers assigned to conduct the focused ethnographies were Bangladeshi. They had academic backgrounds in anthropology, sociology, and public health and were with 2–5 years of qualitative research experience. Moreover, they had 3-weeks of training prior to the ethnographic field research. One researcher belonged to the Marma indigenous community, and shared the linguistic, religious (Buddhist), and cultural background of the Marma community in Roangchari, Bandarban. He also had an understanding of other indigenous communities’ cultures, living in that location. All researchers were trained to be conscious of the power dynamic that is often at play between the respondents and themselves, stemming from differences in social class and demographic characteristics. An awareness of this power dynamic prompts researchers to approach the respondents with maximum sensitivity and respect. Throughout the qualitative phase of this study, all researchers in the field engaged in a reflexive process and continuously questioned their own preconceived notions and assumptions. Researchers took field notes and maintained research integrity. They kept observation notes of the field along with reflection diaries. This ensured that the participants’ points of view were captured in their truest form. All the interview transcripts were triangulated with respective observation notes and reflection diaries. All the qualitative researchers were involved to complete the focused ethnography transcripts of their respective participants, and they were responsible for coding the qualitative data for analysis.

Regular debriefing sessions among researchers were held during data collection to identify challenges and mitigation strategies of data collection, and to discuss the data and its quality. During data analysis, researchers conducted discussion sessions among themselves, triangulated and compared data collected through different methods, a variety of data sources (respondents and study sites). Thus, researchers ruminated on multiple perspectives to interpret data.

Reflections on the Methodological Challenges and Mitigation Strategies

Several methodological challenges were encountered throughout the quantitative (cross-sectional survey) and qualitative (focused ethnography) phases of the study. Challenges during both phases mainly involved building trust with respondents and community members, selecting, and accessing field sites (due to long distance and non-mapped areas), dealing with respondents dropping out, and navigating difficult political situations. Since we conducted the study across the country, covering multiple divisions and regions, field level challenges were heavily influenced by the climate, local geography, and the explanation of validated scales used in the survey.

Tool Development: Language Barriers and Adjusting to the Local Context

We developed a comprehensive questionnaire for the nationwide cross-sectional survey. We included 21 sections (topics) that include i) Socio-economic condition; ii) Family and social network related information; iii) Wealth Index; iv) Daily life: Personal activities; v) Aspirations Index; vi) Use of information Technology; vii) Masculinity Scale; viii) Knowledge on puberty; ix) Knowledge on SRHR (family planning, on HIV/AIDS, STI); x) Gender Equitable Men (GEM) scale; xi) Perceptions, Experiences and practices regarding SRH issues; xii) Knowledge and experience on marriage; xiv) Reproductive coercion; xv) Sexual Harassment; xvi) Risk taking behaviour; xvii) Sexual relation and performance; xviii) Fatherhood; xix) Health service utilization for self and perceptions; xx) Program affiliation; and xxi) Religious activity and beliefs. This questionnaire was developed using validated tools from Demographic Health Survey (DHS),^{1, 2} Technology use, ³ Aspiration Index,^{4,5, 6} Gender Role Conflict/Stress (GRC/S) Scale, ⁷ Gender-Equitable Men (GEM) Scale, ⁸ and the Duke University Religion Index (DUREL). ⁹ For the qualitative study, we utilized separate interview guidelines for individual case studies, focus group discussions, parent’s in-depth interviews, key informant interviews with formal and informal healthcare providers, and community influential persons.

Although these tools were validated, we found that after translation to the local language, some terms needed to be revised to be made culturally appropriate. For instance, during the pilot, we found that some respondents found it difficult to distinguish between questions of the 35-item multi-domain Aspiration Index (AI) developed by Kasser and Ryan (Kasser & Ryan, 1993, 1996). For example, under the ‘Financial Success’ domain, the AI uses two statements: “I want to be a very wealthy person” and “I want to be rich”. Our respondents mostly failed to differentiate between these two statements even though both cannot be interpreted in the same way. Similarly, under the ‘Physical Fitness’ domain, statements like “I want to be physically healthy” and “I want to keep myself healthy and well” were perceived to be similar by the respondents. After consulting these issues with the AI developers, we modified our survey instruments and changed it to a 23-item scale which better aligned with the respondents’ understanding of financial and economic terms.

Evidence on masculinity, sexuality, sexual practices and behaviour, and gender-based violence pertaining to young males in Bangladesh is very limited. This research study initially reviewed and adopted some of the tools used in other countries, mostly in the Americas and Europe. However, we found that these tools needed to be further modified and contextualized. According to the social norms in Bangladesh, discussions around sexuality, sex, contraception, et cetera, are considered taboo. For example, during the pilot, the respondents evaded and seemed embarrassed by questions of sexual relationships and the use of terms like “sex”. As a result, we revised the term to “physical relations” (“sharirik shomporko” in Bengali) from “sexual relations” (“jouno shomporko” in Bengali). Since this study is primarily focused on sensitive SRH topics, we organized the survey tools so that the questionnaire would begin with general information (socio-demographic and socioeconomic data) to allow participants some time to become comfortable and slowly ease into the questionnaire. This was followed by questions on respondents’ daily activities, life aspirations, health-seeking behaviours, and perceptions about SRHR. The questionnaire then gradually steered towards more sensitive topics such as sexual practices, drug use, reproductive coercion, and gender-based violence.

During the development of the study tools (survey questionnaire and qualitative interview guidelines), we consulted with multiple national and international experts, including researchers, academics, practitioners, government, and non-government stakeholders, and incorporated their feedback and comments. Furthermore, all tools were pretested three times, and were modified based on these experiences prior to implementing the study.

Obtaining Authorization from Local Government, Gaining Community Trust and Relocating Field Sites

Due to ethical and practical reasons, we had to obtain permission from key local authorities and representatives of the community to be able to conduct our research study smoothly. Even with permission letters obtained from BRAC University, the Directorate General of Family Planning (DGFP), Upazila Nirbahi Officer (UNO) (head of an administrative region), and metropolitan police (City Corporations), we found it difficult to gain entry and build trust at the village and household level.

At the village level, local authorities expressed concerns as a result of rumours associated with bKash (a mobile financial service company) fraud, terrorism (male youth were thought to be coerced by religious groups into joining the latter’s cause), and a general distrust of outsiders. At the household level, building trust proved to be a time-consuming process. The level of education and temperament of potential male respondents varied. Some qualitative researchers reported that they only truly got close to their respondents during the last week of fieldwork. In addition to the difficulty of building trust between respondents belonging to ethnic minorities and Bangladeshi researchers, the differences in dialect proved to be another major challenge.

Obtaining respondents’ consent and the community’s trust during the survey period was challenging. The team faced several obstacles during the pilot. For instance, some locals became suspicious of the research team (comprising of 25–35-year-old male enumerators), rumours spread that the team intended to recruit the village youth into extremist groups while others claimed that the research team members were child abductors. At that time, the country was seeing a spate of mob attacks on individuals suspected to be child kidnappers.^{10, 11} One of our research teams (of eight members) was chased, confined, and had their equipment confiscated by the locals. They were held there until the team was able to prove that BRAC University and the DGFP were part of the study. The team was requested to leave the area, and the team members had to comply. Following this incident, research teams quickly realized that they had to formally collect approval letters from local governments and law enforcement agencies in the data collection areas. Furthermore, the data collection teams acquired official approval from central and local authorities and verbal approval from local social and political leaders. This was a complex and time-consuming process which led to delays in data collection in several locations. However, this strategy was essential for the smooth continuation of fieldwork.

Given the age profile of data enumerators, community people of different clusters still showed apprehension during initial meetings, as they perceived enumerators as agents of government law enforcement agencies collecting information on young males involved in anti-social behaviour or religious extremism. The research team had to repeatedly clarify the objectives of the study through multiple rounds of discussions, and present documents of authorization from national and local government authorities and law enforcement agencies in addition to confirming the support of BRAC’s field level staff. This was supplemented further by providing hard copies of the survey to local community leaders for review. These measures helped remove apprehension towards the research teams and established transparency to build trust.

There was a concern of bKash fraud (the leading mobile money app in Bangladesh, one of BRAC’s sister concerns) among community people. Fraudsters collected bKash numbers from agents across the country, then called users masking numbers to acquire information (Bangla Tribune, 2019). Since we were associated with BRAC, locals in many clusters misconstrued the data collectors’ intentions and identities as fraudulent. In the Brahmanbaria district, the local people confiscated one of the survey team’s instruments and held them captive in the middle of a haor (a wetland). We received support from the local BRAC NGO official along with the Union Chairman. They helped resolve the situation and convinced the local community members to free the enumerators.

The qualitative researchers conducted one Participatory Rural Appraisal (PRA) and drafted a community map and historical timeline for each of the locations in the four selected sites. The researchers initially invited all village elites, elders, and community leaders to a designated place. The study objectives were explained and information on the village/community was collected which also helped build rapport in the process. Some strategies included reaching young males through their school teachers who introduced them to the researchers, and rapport was built from there. Also, the researchers took the initiative to be familiar and socialize with the respondents during the evenings in open fields or near the hill (in Roangchari) areas where young males of the communities frequently gathered. Even with these approaches, communities in Roangchari continued to express negative attitudes towards one Bengali qualitative researcher in the backdrop of ongoing disputes between indigenous groups and Bengali communities in the region, which caused the highest number of case study drop-outs (six) in the Bandarban area. Therefore, the Bengali researcher with the support of researcher from indigenous community in Bandarban had to convince more study participants and conducted the case studies with those who were motivated and understood the objectives of this study.

Accessibility to Young Males

One of the biggest challenges was reaching young male respondents and completing full interviews in both the quantitative and qualitative data collection periods. As per the survey sampling method, we targeted every fifth household to reach respondents aged 15–24 years. However, the selected demographic presented varying time constraints. Those in the older age bracket (20–24 years) were involved with work and earning an income, whereas the younger respondents (15–20 years) had school and, in some cases, were also involved in work (either the family business or a job). Most males are mobile, and so it was difficult to find them at home, unlike adolescent females who, after school or even with work, return home and usually remain there. Therefore, it was difficult to find male respondents at home, as the time schedule of the interview overlapped with either school or work hours. Many potential respondents who were neither enrolled at school nor held a job were also not available.

During Ramadan, the month of fasting observed by Muslims, many respondents stayed at the mosque saying prayers till late night. During Ramadan there are special prayers (Tarawih) that are recited at night. This resulted in researchers having to make multiple visits to complete interviews, adding to the overall time spent in each cluster. Respondents were dropped from the sample when researchers failed to reach them after two consecutive visits. For the qualitative interviews, when respondents were unavailable during the daytime or evening, only some were willing to participate in the interviews at night. However, due to safety concerns, sometimes it became difficult for researchers to conduct interviews at night, particularly in urban slums and hill tract areas.

Obtaining Consent and Respondents Dropping Out

SRH issues are highly sensitive and treated as taboo in the context of Bangladesh (Amin et al., 2020; Camellia et al., 2020). The majority of SRH related interventions in Bangladesh over the past 10 years have focused on women and girls (Ainul, Bajracharya, et al., 2017). Therefore, SRH issues are largely considered to be “women’s issues”. When researchers approached young males and community people, there was some confusion as they were unable to understand the necessity of such research. They often asked, “Why are you talking with young boys? These (sexual and reproductive health) are women’s issues, not men’s issues.” These perceptions resulted in some scepticism initially, as previous experiences of seeing women and girls in their communities or villages involved in SRH-related surveys or programmatic interventions led them to believe that such topics are only to be discussed with women.

Prior to data collection, informed consent (both verbal and written) was sought from respondents (who were above 18 years old) and parents of respondents (who were under 18 years old). Verbal and written assent was also taken from respondents under 18 years. Some parents were hesitant to let an unknown person interview their son, and the sensitive nature of the survey added to this hesitancy. However, respondents under 18-year-olds easily provided interviews without reservation when they knew that their guardians (mainly parents) allowed them to be interviewed. In some cases, respondents who had previously provided consent discontinued the interview when questions regarding sexual practices were posed. Some also expressed fears of being labelled as religious terrorists and having their data collected. Fears were associated with their responses spreading throughout their community, whereas others feared that their ID data would be extracted by the government, and that they would be labelled as terrorists. As with many other survey-based studies, despite multiple piloting sessions, the survey length was a deterrent to many respondents, as they were not willing to forego other social activities to participate in this survey. For them, there was no motivation to participate. To overcome this problem, the researchers dedicated extra time and engaged with respondents to build rapport and motivated respondents by explaining the context and future implications of the study. For example, researchers shared their own SRH experiences, and explained to respondents that the latter’s participation in this study would go a long way in shedding light on young men’s SRH issues in Bangladesh, a topic that is far too often ignored.

In the sub-set of qualitative ethnographic case studies, some respondents first agreed to participate, but later, they chose not to continue and were unwilling to communicate with the researchers. Researchers were prompted to reschedule multiple times. However, more than five case study respondents dropped out after only two to four sessions as they lost interest or were unwilling to continue or complete those sessions.

Moreover, as with the quantitative survey, there were concerns regarding the privacy of information. Though researchers ensured data confidentiality and research integrity throughout the process, respondents expressed worries about their information becoming public knowledge in their communities. These worries were primarily associated with possible backlash of teasing by their peers and shame from the community. Furthermore, previous respondents who had dropped out tried to discourage other boys in the community from participating as they were afraid that their sexual history and drug use would be revealed in interviews and could spread throughout the community, leading to backlash.

In order to ensure smooth data collection and match the timing and availability of respondents, focused ethnography sessions were divided into six sessions, with the first two sessions dedicated towards ice-breaking and study explanations, and the subsequent four sessions focused on the core SRH objectives. To convince the parents of under-18 respondents, the researchers shared a brief of the whole questionnaire (during the survey) and described sections of the guideline (during the qualitative study) and replied to their queries. We also requested and convinced them to not be present during the interview of their respective male under 18 children, as it could hamper the validity of the information.

However, despite prolonged efforts, 10 respondents dropped out of the case studies after initial agreement to participate, either due to employment opportunities outside of the village or because of discouragement by their peers. For example, a respondent in Sholoshohor, Chattogram had received a 20-day temporary job offer in Dhaka, the capital city, and left after 2 days of the ethnography. Such incidents added to the challenges of finding new respondents and orienting them. Despite all these stumbling blocks, young males, in general, showed enthusiasm to participate. Data collectors and researchers did not find any major differences and challenges regarding the execution of the survey questionnaire or case study interview, and FGDs with under 18 young males compared to older groups. However, at times the researchers had to explain some of the SRHR terminologies to the young respondents so that respondents could answer accurately.

Maintaining Privacy and Confidentiality

Since the survey consisted of questions related to personal relationships, STDs, sexual health behaviour, violence, and drug use, respondents in all districts raised concerns regarding data confidentiality. Family members and neighbours expressed curiosity to listen to the discussions between the researchers and the respondents. To counter these interferences and maintain privacy, enumerators along with respondents dedicated time to find secluded or private spots to carry out interviews. Such areas included quiet places like empty clubs or schools, pond side areas, and paddy fields.

The respondents were very reserved in sharing information on multiple sexual partners, drug use, love affairs, STDs, gender-based violence and other SRH issues. Respondents also expressed concerns regarding the possible spread of personal and SRH information, explaining that it would be detrimental to their reputation if others in their locality were to find out about the details of their sexual history. As mentioned previously, the enumerators shared their own experiences and practices to help respondents open up, and they reassured participants of data confidentiality.

During the ethnographic case studies, researchers lived with the respective communities for a month. Much like the enumerators in the survey, ethnographic researchers chose private places for the interview sessions to avoid any kind of interruption. Full privacy was still difficult to achieve, leading some sessions to be conducted on fields or riverbanks. The researchers refrained from discussing the study with participants except in private spaces and avoided conducting interviews in public.

Explanation of SRHR Topics and Associated Terminologies

Since this study involved all aspects of SRHR, including family planning, sexual habits, and STIs/STDs, researchers noted that during discussions about STIs/STDs, respondents from every research site had little to no knowledge of other STIs other than AIDS. Researchers also faced difficulty in explaining the general definition of STI and the different types of STIs (e.g., herpes, gonorrhoea, hepatitis B, syphilis, and genital warts).

Difficulties were compounded due to the usage of different languages and dialects across study sites. For example, the region of Roangchori, Bandarban is split into ethnic groups which together form six to seven indigenous groups. Each indigenous group has its own dialect which has significant variations from dialects of other indigenous groups. Thus, it was difficult for researchers to determine the appropriate terms related to SRH in the different dialects. For example, the term for masturbation was unknown to some respondents. To overcome such language barriers, the researchers who conducted the ethnographies used stories of everyday life to figure out the local terminologies. This strategy was used to ascertain the terms for relationships, sex and wet dreams.

Introducing Scale-based Questions

The survey employed scale-based questions with scale order ranks. It was very challenging for the enumerators to explain these questions and collect data accordingly. In particular, scaling responses using Likert psychometric scales, the Gender Role Conflict scale (GRC/S) and the Gender Equitable Men’s scale (GEM) not only had statements related to sexuality and violence, but were also ordered to capture the intensity of feelings towards a given item. This required extended periods of explanation through examples to avoid central tendency bias in selection (Landy & Conte, 2009) and reduce serial position effects on answers (Mcleod, 2008). This added, on average, an additional 20 min to the estimated survey time of 50–60 min.

Triangulation: Establishing trustworthiness and credibility of Data

During qualitative data collection and analysis, we used several approaches for data triangulation to increase the validity and credibility of the findings. For triangulation, we used different methods including case studies, reflection diaries and observation notes, focus group discussions, in-depth Interviews, and key-informant interviews. The researchers also discussed and consulted amongst themselves regularly and used multiple perspectives to interpret the data. A weekly debriefing meeting was held among researchers with the other research teams in different locations during data collection to discuss the findings. During focused ethnography data collection, researchers stayed in the field sites for one and-a-half months. There, they had informal discussions with local people, which became a secondary source for cross-checking data provided by the respondents. This enabled researchers to get a sense of underreported data and exaggeration. For example, a respondent in Chittagong detailed his relationship with his girlfriend but kept denying being engaged in any sort of sexual activity with her, despite reporting taking trips with his girlfriend. Also, in the same region, a 16-year-old boy claimed to have lost count of his sexual encounters (starting from the age of 12), stating “How can one keep track of these things?” He also reported having visited multiple brothels in the past three to fouryears. Towards the end of the focused ethnography, one 19-year-old unmarried male respondent from Satkhira was being taunted by his friends. They were taunting him by repeatedly shouting out another adolescent boy’s name. Upon asking these boys why they were doing so, the researchers were told that the respondent had sexual intercourse with the boy in question, which the other boys claimed to witness. The respondent had mentioned having no sexual relationships with males before. However, when he was probed again, he confessed to being in a same-sex relationship, along with having other sexual encounters with the opposite sex.

Conclusion

This reflective account of fieldwork experiences of a nationwide survey and focused ethnography on the SRHR of young males in Bangladesh provides insight into the methodological challenges that often arise in projects involving sensitive topics such as sex and sexuality. Overall, participants in both the quantitative and qualitative phases of the study showed a willingness to participate, often expressing interest in advancing research/services targeting men. This finding contrasts with other SRHR-related research experiences according to which men were found to be less interested in supporting research and development program activities compared to women (Mohajer, 2019). However, the willingness to participate was not equal across all regions, as per our observations during the focused ethnography. Respondents belonging to ethnic minority groups in Bandarban showed reluctance to participate for reasons related to mistrust of outsiders (in this case a Bengali researcher) with its roots in the country’s political history.

Several young males discouraged their peers not to talk with researchers. Here, young men’s ignorance towards researchers and a tendency to refuse to cooperate (either due to time constraints or lack of interest) can be interpreted as a way to demonstrate power (Toerien & Durrheim, 2001). Since researchers are outsiders and respondents do not directly benefit from participating in research projects, young men do not give much thought to the ultimate objectives of participation in such research (Williams & Heikes, 1993).

The researchers encountered challenges in obtaining consent and conducting full interviews as respondents could not always dedicate the time needed to complete the interviews due to other personal, social, and economic commitments. Selecting case study respondents for the qualitative phase was also a challenge. Potential participants were recruited though mutual friends, peers, and young local gatekeepers. This is a very common and useful technique of recruitment cited by other researchers (Oliffe & Mroz, 2005; Ryan et al., 2019).

Young males tend to avoid sharing their emotions and experiences on sensitive issues such as sexual feelings and practices, drug use, and gender-based violence (Robb, 2020). Respondents perceived open display of emotions and reporting of sex practices as a poor reflection on their personality. Other studies also reported similar challenges, as men tend to treat illnesses and emotions as private matters (Lee & Owens, 2002). Young males are often uncomfortable expressing their vulnerabilities (O’Neil, 2013), and are relatively less expressive when describing their experiences and memories compared to women (Bottaro, 2018). Researchers may overcome these challenges by establishing trust through good rapport-building with respondents, particularly by conducting one-to-one sessions that maintain privacy and confidentiality (Frosh et al., 2001), which was practiced in this study.

Furthermore, evidence shows that the sex of the researcher significantly influences researcher-participant relationship dynamics and the information shared by the respondents, particularly when sensitive topics are involved (Hutchinson et al., 2002). Since the respondents of this study were young men, male researchers and data enumerators of a similar age bracket was deployed to collect data. Therefore, respondents felt more inclined and comfortable to share sensitive information, which led to a reciprocal relationship between the researcher and participant during interviews (Krueger & Casey, 2014).

Areas that warrant careful consideration for future studies are information pertaining to private matters and emotions of men. According to our experiences, taking extended time for rapport-building can lead to more efficient data collection procedures, especially when ethnography is involved. Allocating ample amount of time to carry out prolonged sessions for rapport-building with respondents is a core component of efficient data collection. In the qualitative phase of our study, PRA methods and 2 days of rapport-building were supplemented by numerous informal discussions which helped us understand the nature and norms of the community, and provided an opportunity to introduce and explain the study objectives. With regards to the quantitative phase, in addition to a focused rapport-building process, special consideration should be given to scale questions, keeping in mind the sensitive nature of SRHR topics, and the scale order ranking to be done by the participants.

Regardless of the methods used, the researchers of this study concur with the statements “Data collection is multi-method and often continues until saturation” (Leary, 2017, p. 120) and that “It is difficult to anticipate all possible issues before undertaking research; there are always unanticipated issues in the field” (Lu & Gatua, 2014, p. 13). Therefore, a flexible approach should always be considered, without compromising the standards of research, especially when conducting large-scale studies on sensitive topics such as SRHR. Importantly, particular attention ought to be paid to the local social context, norms, and mores, so as to ensure that researchers engage in respectful conduct and language and use culturally appropriate tools.

There is a need for increased reflective accounts of experiences in quantitative, qualitative, and mixed methods research in the field of SRHR, particularly in the context of LMICs where conducting such research can be especially difficult. The literature on the methodological obstacles encountered in SRHR-related research in LMICs is scant. Detailed accounts of challenges and mitigation strategies adopted by researchers exploring oft-overlooked SRHR topics can serve as a useful resource for less experienced researchers. Knowledge exchange of such kind can go a long way in increasing researchers’ familiarity with expected challenges and enhancing their preparedness.