Our analysis included 5,196 participants from Honduras, 2,436 from El Salvador, 2,408 from Côte d’Ivoire, and 8,568 from Lesotho. The weighted sample size ranged from 491,639 in Lesotho to 5,213,050 in Côte d’Ivoire. Across all four countries, male and female participants were evenly distributed (Table 1). There was a range of variability with regard to demographic variable prevalence in the four countries. In total, and varied by country, 14.8%– 28.4% of participants were married or lived with someone, 40.4%–56.8% had ever had sex, 10.8%–45.8% were orphans, and 33.2%–55.3% had completed primary school or less. Depending on the country, 21.6%–33.0% of participants reported witnessing violence in the community, and the prevalence of witnessing violence in the home ranged from 16.4% to 47.9%. The prevalence of lifetime EV ranged from 12.1% to 26.3%, whereas lifetime PV exposure ranged from 23.4% to 68.2%, and lifetime SV exposure ranged from 8.3% to 25.8%. The proportion of youth who indicated they were afraid someone might hear their answers was low for these four countries, ranging from 3.0% to 17.1%. In total, 2.0%– 17.5% of youth found it upsetting or stressful to answer the survey questions. More than 90% of youth across the four countries found it worthwhile to participate in the survey, ranging from 93.3% to 98.6% (all proportions, by country, are shown in Table 1).

Table 2 includes the unadjusted weighted percentages and adjusted odds ratios of fearing that someone might hear the answers, stratified by demographic characteristics and violence experiences. Female participants in Honduras, aOR = 1.6, 95% CI [1.2, 2.1] and Lesotho, aOR = 1.9, 95% CI [1.1, 3.2] had higher odds of being afraid of being overheard compared to male participants. In Honduras, youth aged 13–17 years had 1.5 higher odds, of being afraid of being overheard compared with those aged 18–24 years. In Honduras, aOR = 1.6, 95% CI [1.1, 2.2], and Côte d’Ivoire, aOR = 2.1, 95% CI [1.1, 3.9], not being married or living with someone was associated with higher odds of the fear of someone hearing one’s survey answers. In Honduras, aOR = 1.4, 95% CI [1.0, 1.9], and Côte d’Ivoire, aOR = 1.7, 95% CI [1.2, 2.4], ever having had sex was associated with higher odds of being afraid of being overheard. Neither orphan status nor witnessing violence in the community was significantly associated with the fear of being overheard in any country. In Honduras, youth who completed primary school or less had 1.5 higher odds, 95% CI [1.2, 1.9], of being afraid of being overheard. In Côte d’Ivoire, youth who witnessed violence in the home had 1.5 higher odds, 95% CI [1.1, 2.0], of being afraid of being overheard than those who did not witness violence in the home. Across all countries, experiencing EV was associated with higher odds of being afraid that someone would hear survey answers, Honduras: aOR = 1.8, 95% CI [1.4, 2.4]; El Salvador: aOR = 3.7, 95% CI [1.5, 9.6], Côte d’Ivoire: aOR = 1.7, 95% CI [1.2, 2.4]; Lesotho: aOR = 2.1, 95% CI [1.4, 3.1]. Similarly, experiencing SV was associated with higher odds of being afraid that someone would hear one’s survey answers across all countries, Honduras: aOR = 2.6, 95% CI [2.1, 3.3]; El Salvador: aOR = 2.2, 95% CI [1.2, 4.1]; Côte d’Ivoire: aOR = 2.1, 95% CI [1.4, 3.0]; Lesotho: aOR = 1.6, 95% CI [1.1, 2.2]. In Honduras, experiencing PV was associated with 1.4 higher odds, 95% CI [1.1, 1.8], of being afraid someone would overhear.

Table 3 shows the unadjusted weighted percentages and adjusted odds ratios for finding participation upsetting or stressful, by demographic characteristics and violence experiences. Higher odds of finding participation upsetting or stressful were found among female participants in Lesotho, aOR = 1.6, 95% CI [1.0, 2.4]; youth who had never been married or cohabitated in Honduras, aOR = 1.7, 95% CI [1.1, 2.7], and Lesotho, aOR = 1.5, 95% CI [1.1, 2.1]; and orphans in El Salvador, aOR = 3.8, 95% CI [1.6, 8.8]. In each country, some experiences of violence were significantly associated with finding participation upsetting or stressful. In Honduras and El Salvador, youth who experienced EV, Honduras: aOR = 2.3, 95% CI [1.5, 3.5], El Salvador: aOR = 2.7, 95% CI [1.2, 6.3], and SV, Honduras: aOR = 2.0, 95% CI [1.3, 3.0], El Salvador: aOR = 3.0, 95% CI [1.4, 6.5], had higher odds of finding participation upsetting or stressful than those who did not experience EV or SV, respectively. In Côte d’Ivoire and Lesotho, youth who witnessed violence in the home, Côte d’Ivoire: aOR = 1.4, 95% CI [1.0, 1.8], Lesotho: aOR = 1.5, 95% CI [1.1, 1.9]; experienced EV, Côte d’Ivoire: aOR = 1.7, 95% CI [1.1, 2.6], Lesotho: aOR = 3.4, 95% CI [2.6, 4.5]; experienced PV, Côte d’Ivoire: aOR = 1.8, 95% CI [1.0, 3.0], Lesotho: aOR = 1.5, 95% CI [1.1, 1.9]; and experienced SV, Côte d’Ivoire, aOR = 3.3, 95% CI [2.3, 4.8], Lesotho: aOR = 2.3, 95% CI [1.8, 3.0], had higher odds of finding survey participation upsetting or stressful.

We assessed the reported reactions to being asked difficult questions about violence exposures in a sample of 18,608 individuals aged 13–24 years across four countries (i.e., two in Africa and two in Central America) as part of country-specific survey interviews about experiences of violence during childhood and adolescence. Several demographic characteristics were associated with distress related to survey participation; however, patterns of findings were inconsistent across countries. Across all countries, more than 90% of children and youth found survey participation worthwhile. Fewer than 18% of children and youth in any country, with fewer than 4% of participants in Honduras and El Salvador and approximately 8% in Lesotho, reported being upset or stressed by being asked questions about their exposures to violence or expressed fear that anyone might hear their answers during the interviews.

VACS are the only global nationally representative surveys that collect comprehensive data on violence among children and youth. These surveys use a rigorous methodology to ensure data quality while preserving confidentiality, protecting survey staff and participants, and providing response plans to give services to participants in need of assistance. These findings highlight the value of VACS data and suggest that negative reactions were relatively uncommon in three of the four countries examined. The findings also show a high degree of consistency across countries, suggesting that these findings are neither contextually nor culturally specific. This study extends and further affirms Deprince and colleagues’ (2006) finding that these questions are not harmful to participants; rather, participants value survey participation. Similar to other studies (McClinton Appollis et al., 2015; Sikweyiya et al., 2021), the present findings demonstrate that negative reactions were relatively uncommon, and when they occurred, there was no evidence of harm or a negative impact of participation. This may suggest that answering questions about experiences of violence that a respondent associates with feeling upset or stressed are not likely to result in long-term harm from VACS participation.

Reactions to questions about experiences of violence may be modulated by the types of violence experienced and the context in which the interviews take place. EV and SV were associated with more negative responses to survey participation, albeit with relatively small effects. Individuals who experienced SV expressed more fear of someone hearing their answers and more feelings of being stressed or upset by the questions relative to those who were not exposed to SV. Experiences of SV can carry more stigma (Kennedy & Prock, 2018) among survivors, including victim-blaming messages as well as specific stigmatizing reactions from others. EV exposure was also associated with more fear of being heard and reported distress regarding survey questions and answers relative to no EV exposure. Although most youths did not find participation stressful, these findings suggest that many participants who disclosed their experiences of violence to the interviewer retained some privacy concerns and experienced some degree of stress. Geographic and cultural differences can also account for the differences between the surveys.

The context of the interview is also key. Documenting the perceptions and reactions of interviewees during interviews that address violence experiences can inform researchers and practitioners about how best to measure these experiences ethically while minimizing harm. It can also create opportunities for service referrals for participants who may need them. Previous literature (Cater & Øverlien, 2014; Morris et al., 2012; WHO, 2007) has highlighted the importance of conducting interviews with children and youth in safe environments, following clear consent procedures, and communicating about the questions being asked. Asking about experiences of violence and other sensitive topics may, nonetheless, lead to distress among respondents. A study conducted in the United States found that about 23% of 10-15-year-old survey respondents reported feeling upset by questions about violence (Ybarra et al., 2009). Another U.S. study among 13-18-year-olds found that 30% reported some level of distress when asked questions about violence (Langhinrichsen-Rohling et al., 2006). In other settings, a recent school-based study in Kenya used mixed methods to assess children’s perceptions of being screened about SV (Undie & Mak’anyengo, 2020). The authors found an overall positive reaction among children, school authorities, and parents with regard to screening for SV victimization and observed that among participants, screening tool use was associated with improved communication with and disclosure to parents or school authorities. Another survey of primary school children in Uganda qualitatively assessed the perceptions of children about being asked about their experiences with violence (Devries et al., 2015). The findings demonstrated that interviewees felt more at ease responding when they perceived an environment of trust with interviewers. Assurances of confidentiality contributed to decisions to disclose their experiences and their expectations of help; feeling hopeful that something about the violence they were experiencing would change also contributed to disclosure.

Asking these questions could be useful for prioritizing responses. Ybarra et al. (2009) and McClinton Appollis et al. (2015) suggest that responses to these questions can be used to identify participants who may benefit from additional follow-up. Although more evidence is available from studies that have assessed intimate partner violence among adult populations (Clark et al., 2012; Garcia-Moreno et al., 2005), evidence examining children seems to highlight similar findings. For example, studies among adults and children suggest that the benefits of asking these sensitive questions outweigh the negative consequences of not asking them, including stimulating participants’ feelings of distress or upset, directly causing harm, losing participants, concerns regarding whether it is ethical to disclose stigmatizing information (Becker-Blease & Freyd, 2006; Newman & Kaloupek, 2004; Rinehart et al., 2017), or the fact that some respondents will experience distress during the process of being interviewed (Cromer et al., 2006; Decker et al., 2011). In a systematic review of 46 studies, Jorm and colleagues (2007) concluded that even when participants experience distress, it often dissipates quickly, and their findings suggest that participation in research does not pose more than minimal risk of emotional harm.

The present report highlights characteristics of the participants who reported distress or concerns about disclosure of their experiences of violence. This information can be used to train interviewers and design response strategies aimed at identifying children at higher risk or children who might need additional support in the context of interviews. The proportion of individuals who reported distress was similar across most countries included in this study and is consistent with findings reported in other settings (Ybarra et al., 2009). The higher prevalence of distress in Cote d’Ivoire warrants further investigation.

The present study had several limitations, including a reliance on cross-sectional data, which precludes us from establishing temporal associations. In addition, although the survey was lengthy, only a few questions address respondents’ perceptions of the survey experience, and there are potentially unmeasured social variables that may have contributed to a higher prevalence of violence among certain groups. Qualitative assessments could provide further reasons for the given responses. We cannot determine which specific cultural characteristics might exist that account for the potential observed differences by country. Finally, VACS survey responses are self-reported; thus, they are subject to recall biases and the potential of nondisclosure due to social desirability or fear. Despite these limitations, these findings provide useful information from a large sample of adolescents and young adults from different regions.

This study has relevant implications for practice, including providing researchers with information about elements to include in ethical protocols for violence research and when training interviewers to collect violence-related data. These findings also suggest the need to ensure referrals are available and offered to participants as needed. Second, the findings highlight the importance of ensuring that violence surveys are conducted in an environment of safety and trust, where privacy is ensured and respected. Finally, the VACS surveys invest a considerable amount of time in training in theory and practice on how to properly conduct surveys addressing these sensitive topics. Such training is essential and can be beneficial regarding not only the collection of high-quality data but also in providing participants with an avenue for addressing their experiences and offering assistance if they wish to seek help.