“Children Are Like Vuvuzelas Always Ready to Blow”: Exploring How to Engage Young Children in Violence Research

Abstract

Children’s participation and inclusion in violence research, particularly in low- and middle-income country (LMIC) contexts, is scant and not well understood. To assess how young children can be engaged in violence research, 4- to 7-year-old children were recruited into our pilot study in a rural area of South Africa. Six interviewers, recruited from the community, were trained to complete cognitive interviews (n = 24), interviewer-administered questionnaires (n = 21), and qualitative interviews (n = 18) with young children. Three focus group discussions (FGDs) were conducted with interviewers. Findings from FGDs and assessment of interview performance highlighted that young children could feasibly and meaningfully appraise violence they experience and articulate their view in a research context. Art- and play-based approaches offered participants an easier and developmentally appropriate platform for communication, expression, and engagement, and asking directly about violence was acceptable. The ease with which children participated was determined both by their level of development and the sensitivity of the interviewers; thus, intensive training and mentorship were required over time to assist interviewers in understanding child development and mental health and increase recognition of these issues and their presentation. Interviewers critically engaged with personal values regarding children’s rights and voice in research, reflecting that some of the stories were difficult to listen to. They were able to use and value novel methods to facilitate the ethical involvement of young children to yield rich data. Without young children’s involvement and dynamic participation in violence research, the field will not have the evidence to build best practices, respond appropriately to the needs of this vulnerable population, and interrupt the intergenerational transmission of violence that develops in these formative years. Our study adds to the burgeoning evidence that young participants are vital to the research process and are valuable active contributors to understanding violence in LMICs.

Introduction

Almost a third of children in low- and middle-income countries (LMICs) have been exposed to intimate partner violence (IPV), with past year prevalence at 35% (Kieselbach et al., 2022). Data from Multiple Indicator Cluster Surveys (country-level household surveys developed by the United Nations Children’s Fund to provide internationally comparable, statistically rigorous data on the situation of children and women) show that children in developing countries experience high rates of physical and emotional abuse from parents and caregivers, largely on account of positive attitudes toward corporal punishment (Akmatov, 2011). In a nationally representative sample of 5631 15- to 17-year-old children in South Africa, levels of emotional, physical, and sexual abuse were 13, 18, and 12%, respectively (Ward et al., 2018). In other more localized research in two South African provinces with a sample of 3514 10- to 17-year-olds, the prevalence was 56.3% for lifetime physical abuse (18.2% past-year incidence), 35.5% for lifetime emotional abuse (12.1% incidence), and 9% for lifetime sexual abuse (5.3% incidence) (Meinck et al., 2016). Although the prevalence and consequences of exposure to violence in young children are well documented, methods used and manner of engaging young children in violence research, especially in LMICs, are not well reported.

Caregivers and professionals offer a certain understanding of children’s needs from their experiences of violence; however, they may not be fully aware of, or may severely underestimate both children’s exposure to violence and what this exposure means for them (Noble-Carr et al., 2021). Evidence supports violence against young children is typically perpetrated by those in the home (Devries et al., 2018); thus, in cases where caregivers are themselves perpetrators of violence, their reports would predictably be less reliable, and obtaining consent for their child to participate in research may present a conflict of interest. In situations where non-offending caregivers and their children are victims of domestic violence, it is typical that what adult victims perceive as the most frightening experience of violence for their children may not be accurate (Woollett et al., 2020). It is clear that children would be best placed to engage in research regarding the violence they experience; yet, there is very little participatory research with young children who experience violence and are under the age of 10 (Schneider et al., 2015).

Violence research with young children is marred by concerns about participant safety, the potential for distress, and young children’s suspected inability to understand and meaningfully engage in research (Ybarra et al., 2009). Of particular concern is that the manner of asking questions, such as surveys or questionnaires, may have a tone that is misunderstood or perceived as accusatory, or there might be a misunderstanding of items on account of children’s age (Ybarra et al., 2009). Evidence shows in older children and adolescents that the benefits of participating in research on violence (i.e., experiencing positive emotions, skill acquisition, improved self-efficacy, and interpersonal relationships) outweigh risks, such as negative emotions, apprehension around confidentiality and privacy, and participation inconvenience (Neelakantan et al., 2022; Woollett et al., 2017).

All evidence presented on child appraisal of violence is from high-income settings. In a study evaluating preschool children’s attitudes and beliefs about IPV, Howell (2011) found that 92 children aged 4 to 6 years were able to respond consistently (Cronbach’s α = .71) with internal reliability on par with that established in a study of older, school-aged children (181 children aged 6–12 years) on measures of attitudes and beliefs about violence (Cronbach’s α = .71) (Graham-Bermann et al., 2007). Although preschool children’s ability to respond consistently on measures of their internal functioning has been questioned, the authors’ finding was consistent with other research indicating preschool children can reliably report their reactions to conflict (Ablow et al., 2009; Miller et al., 2012). Furthermore, this study found that preschool children’s reliability did not correspond with their comfort in answering questions, that is, children’s level of comfort in the interview was not significantly related to their attitude or belief about violence. In addition, a recent feasibility study conducted with 6- to 11-year-old children in the United Kingdom found that younger children can contribute their opinions on sensitive topics such as violence in ways that are measurable, replicable, and reliable (Barter et al., 2022). Cognitive appraisal processes and beliefs about violence offer one of many possible entry points to help children develop adaptive, healthy strategies for meaning-making and processing their experiences with violence (Evans et al., 2022); engagement in research could further facilitate this.

Concerns also weigh heavily on research ethics committees (RECs) where potential danger inherent in violent research needs to be balanced against an appreciation of the significance of including children (Morris et al., 2012). Apprehension of children’s welfare leads to extensive delays in securing ethical approval and funding. Evidence suggests RECs may assess the risk, at least on prevalence surveys of child maltreatment studies, using higher than normal thresholds based on opinions of possible high distress to participants (Mathews et al., 2022). In a recent review, the analysis demonstrated that participant distress was infrequent and transitory and that legal liability does not extend to emotional distress (Mathews et al., 2022). Instead, good clinical practice principles can be utilized to meet and fulfill ethical requirements for participants. Increasingly, there is evidence to demonstrate that while being asked about traumatic experiences may cause some upset, particularly for younger children, on balance, inquiries into prior episodes of childhood victimization are usually well tolerated by children and adults who experienced violence in childhood (Kyegombe et al., 2019). Children may also appreciate the opportunity to tell their stories and be heard (Radford et al., 2017), which is arguably more “fair” and “inclusive” in terms of philosophical principles of ethics (Bodén, 2021).

The almost worldwide ratification of the UN Convention on the Rights of the Child during the 1990s represents a shift from viewing children as passive objects subordinate to their caregivers to human beings with legal rights (Cater & Øverlien, 2014). From a rights-based perspective, children are subjects of their rights and duty-bearers are responsible for ensuring the realization of those rights (Collins et al., 2021). Tensions between children’s protection and their developing autonomy need to be carefully considered. Adults, who wield more power and privilege in relation to children, need to judiciously deliberate their positionality and the right of children to be heard, especially regarding matters that affect them (children) personally such as violence, in the home, in the community, and at school. Indeed, issues of childism, that is, the phenomenon of prejudice against children, are indicated here. Children are frequently marginalized on account of unchecked societal injustice against them; policy decisions made by adults without consideration for children can have detrimental effects on children’s well-being, for example, global lockdowns instituted during the COVID-19 pandemic have had detrimental longstanding effects on children’s education, food security, and violence against them, particularly in LMICs (Adami & Dineen, 2021). Scholars have urged researchers and policymakers to use the necessary critical lens of childism to deconstruct adultism (or the view that favors adults) and reconstruct more age-inclusive “scholarly and social imaginations” that do not deem a child’s perspective as inferior to an adult’s (Wall, 2022, p. 1) and heed Freire’s (2000) reminder that it is often the oppressed in society that is most silenced.

From our pilot study, we were able to ascertain that young children could be engaged in violence research (Franchino-Olsen et al., 2023; Meinck et al., 2023). The current paper aimed to investigate how young children could meaningfully engage in violence research (i.e., could they answer questions of personal violence experience, were the questions appropriate, did they understand them, and was it upsetting to talk about), what methodologies might be well suited to our objectives, and what was needed on the part of fieldworkers, with scant experience which is typical in our context, to ensure participants could participate effectively. This study also fills a gap in the current understanding of this issue in LMICs.

Method

Background on Pilot Study

In preparation for a longitudinal multigenerational cohort study on mechanisms that underlie the intergenerational transmission of violence (Meinck et al., 2023), a pilot study was conducted in rural Mpumalanga in South Africa from July to October 2021. The study consisted of a convenience sample drawn from residents in the area where the main study was to be undertaken; the pilot study recruited young adults aged 22 to 30 years, children of young adults aged 4 to 7 years, and caregivers of the young adults (Franchino-Olsen et al., 2023). Ethical approval was granted by the University of Edinburgh School of Social and Political Science Research Ethics Committee (264227), the University of the Witwatersrand Human Research Ethics Committee (M190949), and the North-West University Health Research Ethics Committee (NWU-00329-20-A1). Further ethical approval was granted by the Mpumalanga Department of Health (MP-202012-003). Full ethical approval by all ethics boards took 26 months. For more information on the recruitment strategy and measures used, please see Franchino-Olsen et al.’s study (2023).

Participants

Children

First 24 children participated in cognitive interviews, of these, 21 children went on to complete the quantitative interview, and then 18 completed the qualitative interview. Around half the participants from the quantitative interview also completed the qualitative interview. In all, 12 were 4 years old, five were 5 years old, five were 6 years old, and two were 7 years old. In total, 14 participants were girls and 10 were boys. In all, 18 participants were in either a creche or attending school.

Interviewers

Interviewers came from communities where the research was conducted and were believed to have more insight into issues in the community and language ability that could assist with the collection of high-quality data. In research on violence, disclosure to local interviewers is frequently higher than to professional interviewers (Fraga, 2016).

We hired six interviewers (five women), four had completed high school and two had a tertiary qualification. The group of women were aged 48, 42, 31, 29, and 28 years and the man was 28 years. All displayed a thoughtful, non-judgmental attitude and good communication skills. One interviewer had prior experience in violence research; none had conducted research with young children before, although most had children of their own (two younger women did not have children of their own but they had caregiving responsibilities for their siblings). The interviewing process purposively screened for beliefs such as harmful gender stereotypes, and those hired did not appear to condone these.

Interviewer Training and Process

There was substantial investment in interviewer training. This is strongly recommended for violence research (Fraga, 2016) and is significant to research with children as interviewers need a comprehensive understanding of child development, mental health, and outcomes of violence exposure on children. There was significant disclosure of interviewers’ childhood maltreatment histories that needed containment and reflection in training and ongoing mentorship. Disclosure typically occurs when engaged in the experiential practice of the arts-based methods. Some interviewers also disclosed harmful parenting practices in their own lives that required careful attention and discussion to ensure the purpose of the research was fully understood and engaged with. The training focused on balancing the requirements of the job, and the emotional impact of such research. It can be difficult to contain feelings associated with the disclosure of violence and abuse in young children which can lead to mental health challenges such as vicarious trauma in interviewers (Williamson et al., 2020) requiring additional support and recognition. Emphasis was placed on self-care and its practice both in the training and beyond once fieldwork commenced.

Interviewers practiced instruments and processes thoroughly in the training and completed mock interviews, a valuable adult learning method to implement in training child interviewers (Powell et al., 2022).

Consent Procedures

This study reports only on child participation; however, the pilot study recruited young adults and their caregivers as well. Primary caregivers were interviewed first and then gave written consent for their children to participate, enabling them to understand the aims of the study and the kinds of questions being sought to answer. Consent was only obtained from one identified primary caregiver per child. All caregivers gave consent, although some were hesitant and had questions about the length of the questionnaire that needed explanation (i.e., child-friendly and much shorter than the adult questionnaires). Discussions of mandatory reporting occurred with all participants and consenting adults were agreed. Children were invited to be interviewed at their homes, either inside or outside depending on what felt most comfortable to them and where auditory privacy could be assured. After ensuring the child knew that their parent/caregiver had consented to their participation, they were assented to participate. Mandated reporting was also explained in child-friendly terms to ensure child participants were aware of when reporting might be required.

Time was taken to build rapport with children to guarantee a comfort level was reached. This was facilitated by playing an emotional literacy and expression charades game with the added aim of ensuring a clear understanding of feeling states for the mental health components of the questionnaire. Playing with Play-Doh before using it for the questionnaire task and enjoying a squiggle drawing game before engaging in the kinetic family drawing (KFD) task for the qualitative research were additional ways to facilitate rapport with children. For more information about these arts- and play-based methods, see Woollett et al. (2023). Facilitative methods were available but the invitation to engage and the environment provided for storytelling was the responsibility of the interviewer and a skill to be honed and practiced.

Training developed the interviewers’ understanding that young children, who often feel the need to please adults, may not feel able to stop interviews, and, if they experience extreme discomfort, it may show up in their bodies (being hyperactive and jumpy) and their inability to be present (being distracted and unfocused). In such cases, interviewers were instructed to say, for example, “your body is letting me know that we need a break” and were encouraged to play a game and interrupt the interview to help children regulate and refocus, but then invite continuation in the spirit of “continuous consent” asking children “is it okay to carry on?” In addition, fieldworkers wrote observation notes at the end of each interview with a focus on identifying any kind of distress noticed in the interview, as well as feedback from participants on their experience. These notes were also used by the social worker if a referral was made.

In accordance with other research on vulnerable children (Kyegombe et al., 2019), the questionnaire started with less sensitive and emotionally laden questions and ended on a more positive note; the resilience questions were completed last. More emotionally loaded questions were asked in the middle of the questionnaire.

Ethical Considerations

Given special considerations of researching violence, all portions of the study adhered to the World Health Organization (WHO) ethical guidance on IPV research (WHO, 2001), the United Nations Children’s Fund (UNICEF) guidelines on ethical research involving children (Graham, 2013), and guidelines for the prevention and management of vicarious trauma among researchers of IPV (Sexual Violence Research Initiative [SVRI], 2015). All interviewers completed a full background check and two interviewers were encouraged to attend each child interview to ensure the safety of the child. In addition, the second interviewer was able to engage with other family members to create auditory privacy for the interview. A comprehensive distress protocol was employed using a trauma-informed orientation to research engagement and participant interaction (Mathews et al., 2022). Strategies used to ensure children were provided with a safe environment in which to participate included asking them where to conduct the interview, whether it was safe to talk there, engaging in play, collecting data through arts- and play-based methods, providing reimbursement that included a gift pack (valued at $5) consisting of an activity book, a washcloth, and soap. If children engaged in multiple interviews, they also received stickers and a pen.

In recognition that we were researching a vulnerable population and within a complicated research setting, the study employed a full-time social worker to attend to any mandated reporting requirements and further support participants and their families. The social worker attended all training to understand the duties of the interviewers and the purpose of the study and was integrated into the fieldwork team (Devries et al., 2015a).

Data Collection

Cognitive interviews (n = 24), quantitative interviews (n = 21), and qualitative interviews (n = 18) were completed with young children. Cognitive interviews involved testing of children’s understanding of questionnaire items (measures included different types of violence, child abuse and neglect, bullying, and mental health), an assessment of question appropriateness, and evaluating language use and translation fit. Quantitative interviews were interviewer-led, 1 hour in length, focused on children’s relationships, different types of violence exposure, mental health, and resilience, and included arts-and play-based elements. The measures tested in the cognitive interviews were utilized in the surveys. These were completed on Android tablets with the use of visual prompts and tactile aids. Qualitative interviews used drawing methods, including a squiggle drawing to warm up and the KFD (Burns, 1972) used a semi-structured interview schedule focused on perceived interpersonal relationship quality between family members. The qualitative interview was completed after the quantitative interview and at another sitting, usually lasting 40 minutes. Paper and crayons, pencils, and so on were offered for the tasks. Interviewers completed field notes after these interviews to reflect on the process and how well the child engaged in the research. All children were invited to give assent after their caregivers had been interviewed and had provided consent for their child to be interviewed. Consequently, the topics that would be covered in the interview with the children were clear to caregivers providing consent. For more information about procedures, please see the pilot study findings paper (Franchino-Olsen et al., 2023).

Three FGDs with six interviewers were conducted after the pilot study was completed over a period of 4 months to understand experiences of conducting interviews with children and reflect on interviewer concerns and lessons learned. The FGD guides were developed by the investigative team (five members); a team-based approach that helped increase insight and validity of results. These discussions were led in person by the project manager of the pilot study, who is a qualitative researcher and were supported online in real time by the investigative team. Additional questions were asked by investigators during the discussions that took approximately 2 hours, were audio recorded, and transcribed.

Data Analysis

This manuscript reports on qualitative interviews conducted with child participants that were transcribed and translated. First, four investigators reviewed and assessed at least three transcripts each of interviewer-led qualitative interviews, to assess pilot procedures and implementation, giving feedback to interviewers on the utilization of methods, interaction with participants, and quality of data collected. Investigators discussed findings from these transcripts and together with observations from the field and weekly team meeting discussions over a 5-month period, triangulation informed comprehensive discussion guides for the FGDs that followed (Carter et al., 2014). Reflections on key findings after every FGD was completed with the broader investigative team ensured nuanced questions were included in the next round of discussions, confirming that all areas of investigation were being understood and explored. Each FGD covered the various interviews that were conducted in the pilot, namely cognitive, quantitative, and qualitative interviews, and all comprised questions of interviewing young children about violence, how well methods worked in conveying experience, issues of involving young participants in research, what the challenges were for interviewers and what they learned in the process, and so on. The FGD data were analyzed using a rapid analytic approach (Hamilton & Finley, 2019; Vindrola-Padros & Johnson, 2020). This rapid and iterative approach to data collection and analysis was used to develop an understanding of the pilot study, including the acceptability, feasibility, and appropriateness of the study methods and materials, and chosen for its ability to generate targeted research in a timely manner to inform the main study tools and implementation which was imminent (Johnson & Vindrola-Padros, 2017). Study investigators who analyzed the data were embedded in the study with working knowledge of the context and methods, offering a thorough understanding of the complexity of the study implementation as well as the circumstances of participants. The analysis was minimally interpretive with the first author creating an inventory of the data contents or main domains, derived from the discussion guides, condensing and consolidating the data into summaries (Miles, 2018). Three researchers read all transcripts, highlighting relevant quotes that matched the inventory created, and clarified or confirmed summarized key findings collectively. The team met repeatedly to resolve and clarify the areas of disagreement, thus any coding divergence was discussed to reach an agreement.

Findings

How Children can be Engaged in Violence Research

Participants were able to engage in discussion of their experience through drawing, regardless of drawing ability. Drawings were an effective tool for interviewers to use as a prompt and for the participant to talk from.

I think the drawings, it like uhm, those drawings have messages that the children wouldn’t be able to tell you straight, but through the drawings the child is able to tell you. (P5, FGD3)

Children carry a lot of information regarding violence and the only way you can find out is how you ask them, how you engage them, through the drawings they help you get that violence out the child. I believe they give you first hand, raw information. (P2, FGD3)

It appeared appropriate to ask about violence experience directly in all components of the pilot.

I think children are like vuvuzela’s, always ready to blow. I think most of the children we interviewed were already there waiting for someone to listen. Like when a child is very fond of you they tell you everything you will know at the end of the day even things you did not ask. (P2, FGD3)

Some participants did become “activated” by some questions that could be perceived as distress in the interview process. However, it seemed the children were distressed by their living circumstances, not by being asked about them. We did ask child participants about their feelings during the interview and most responses ranged from “I’m happy playing the game with you/drawing with you,” “I am glad you came to talk to me,” to “I’m happy you came to my house.” In addition, interviewer notes completed at the end of the interview reflected that most participants were willing and enthusiastic about being interviewed and spending time with the interviewer and seemed to be proud of their “work” during this time.

Cognitive interviews revealed children were able to answer most items from the questionnaire and were mostly able to repeat these questions back to the interviewer in their own words showing an understanding of the measures used.

Oh, I could remember I was like stunned because, oh, that girl, I forgot her name but she seems to understand each and every question like she could explain in such a way that you can never expect that a 7 year old can answer you in that manner but that girl managed to answer that. I didn’t have like to keep on like explaining or trying to elaborate more on telling her . . . what is this question asking? She was telling me straight “This question is asking me about this and this and this . . .” (P4, FGD1)

The questionnaire worked well on tablets with image prompts to aid comprehension. Tactile visual aids (e.g., three jars of beans/rice filled to full, half empty, and empty levels to correspond with ratings of “always,” “sometimes,” and “never”) helped children to understand abstract concepts and increase engagement (please see Franchino-Olsen et al.’s study, 2023, for examples of this). Children enjoyed participation and were able to take part regardless of age. They were also curious about the tools being used for data collection and seemed fully invested in exploring them, requiring a patient approach on the part of the interviewer. Younger children and those with apparent developmental delays (i.e., those children who seemed to have difficulties with language comprehension and expression or seemed to experience challenges in maintaining focus) were distracted more easily and required more refocusing by the interviewers increasing the duration of interviews.

So I had one, who like when I am busy asking questions, they ask me like “this tablet it’s yours?” so then I say “no it’s from work” then he will continue asking me questions things like “and your phone it’s beautiful!” And “can you borrow me?” Then I borrow him, then we stop the interview, then I borrow him my phone then he touched it and say “I love this phone, can you give it to me?” I said I only have one. So he continues saying “can you borrow me this marker?” I said it is for work. So he said “Yeah, I know, man, I just want to touch the tablet, it's so beautiful.” So I um, gave it to him and he touched it, and we almost stopped for 15 minutes. (P5, FGD2)

Being questioned about violence experience did not deter engagement, even when the violence implicated caregivers. Again, this was a convenience sample, not purposively sampled for violence experience, yet 15 child participants disclosed exposure to violence in the home.

Kids are very honest, I think it’s because they have not reached that stage when they are being told you not supposed to say this, to people who are from outside the family so I think that’s why they are more comfortable to say that “Daddy beats mommy.” They are comfortable about letting it out. They are not afraid. (P3, FGD3)

Parents seemed at times reluctant to consent to their child’s participation after completing their interviews, mostly on account of concerns regarding interview length. When child interviews were explained, all caregivers consented to child participation.

The other parents feel it took very long [their own interview] and will the child cope? So I said to her “yeah actually they will cope because what I’m asking you is not similar to what I’m asking the child. The child one is very short compared to the adult one.” And she was like “Ok. If she wants, talk to her, if she wants and then she will take part. But if she doesn’t want then it’s between the two of you.” (P2, FGD2)

Despite all efforts that were taken to ensure that there was auditory privacy for the interviews with children, some parents found it difficult to allow their children to be out of their earshot creating a challenge for interviewers to ensure confidentiality.

For me, the last one I was interviewing, we moved from another tree and then we go into the house so by then the father used to stand behind the door and you can’t see him . . . And then we had to, to lower our voices and hold the mics in order for him to not hear us because he seems like he wants to hear what we are talking about. (P6, FGD2)

Parents consented to child participation even when they were well informed of mandated reporting requirements and had reservations about social worker involvement in the study, requiring a calm approach by the interviewer to ensure social worker involvement was accepted and useful.

On my side, there was one parent. I told her that they [child participant] were with the social worker . . . So the reaction of the parent was like Oh, will we be arrested here? Will the child tell you that we are punching them? Then I told her that, no, they won’t arrest you, they will talk to you, the social worker will come and talk to you. (P1, FGD2)

Interviewers were able to follow in-depth interview guides to facilitate interviews with participants; however, we did find, through review of their transcripts, that there were some missed opportunities to engage more deeply in questioning and thoughtful probing (see Supplemental Appendix for more detail on this). Lack of experience in facilitating qualitative interviews seemed to lead to a high reliance on the interview guide at the cost of deepening conversations related to our research question.

Impact of Interviewing on Fieldworkers

Interviewers considered it important to ask children directly about their violence exposure to ensure appropriate support could be offered and children’s needs were not minimized on account of caregiver involvement. In all, 11 child participants were referred under mandated reporting requirements to the study social worker who followed these cases up responsively through multiple visits to the family home.

I think it’s a good idea to ask children [about violence experience] because as much as we also asking the adults but the more we ask children the more we are there to assist them because violence comes out and you are able to help the child when they need you. Rather than not asking them and let them stay there and endure violence over and over again. (P2, FGD3)

Interviewers did disclose that listening to some of the stories of the participants was emotionally taxing.

I think I got used to it . . . my first interview it was very emotional. It pained me because it was for the first time interviewing someone whom you had to listen to their painful story. So, by the time went by, I got used to it, but when a participant is now feeling sad I just have to contain and not also cry because this first participant also made me cry. So when I went by on and on so I got used to it and then I was comfortable to contain the participant. (P3, FGD2)

Fieldworkers needed significant training and practice on engaging children in child-friendly ways as this was not necessarily implicitly understood, nor generally engaged in at the community level.

Kids are looking for someone to talk to. Especially at home nobody wants to listen to them; the skill you taught us at the training that you need to be at the same level it makes them comfortable to speak to us and they will tell you everything that is happening around the family. (P6, FGD3)

A great responsibility was placed on interviewers to engage playfully with child participants and for some, this was a difficult aspect to master.

On my side the training helped me more because I wasn’t able to show someone my angry and sad face, it was difficult but then through the training I have learnt and I was enjoying when I was doing it with the child. (P5, FGD3)

Interviewers shared that although some of the skills for this research were difficult to learn, they were beneficial, even being utilized in interviewer personal relationships. Probing skills seemed especially novel and helpful in improving understanding of communication within interpersonal relationships.

I: Okay P, you said the drawings, in the beginning it was very difficult for you to understand the concept and the experience of probing has also stuck with you and you’ve used it in your personal life. So, I’d like to know how has that impacted you in your personal life when you use probing?

P: Sometimes they [family members] say that I am playing with them, I am making them stupid [conscious] but then it helps more like I am able to understand the real thing behind whatever the person is saying. (P5, FGD3)

In addition, some interviewers shared that they learned that some views they held may be inappropriate, raising awareness of and shifting their perspectives of these but also heightening recognition of negative social norms within their personal relationships that required exploration and processing.

For me, it’s that thing that [fellow interviewer] was talking about, that touching part, because I could remember back then in my family of origin, because my sister was having her daughter and my dad used to like touch my sister’s daughter’s breasts and she seemed so happy and that day when [fellow interviewer shared her discomfort of this “game”] I said “oh, what’s this? why did my father do that?” Things like this it is not working. So it’s helped me to digest that some kinds of things they are not good if it happens and you don’t have to take them, even like they, it’s cultural you know? We are not allowed to do that because it can lead that person to go the other direction the other times so I have learned that. (P4, FGD2)

Discussion

Research on violence with children as young as 4 years old is possible. Young children appear to feasibly and meaningfully appraise the violence they experience and express their view of this in a research context. Art- and play-based methods offered participants an easier and developmentally appropriate platform for communication increasing participant ability to be more self-reflexive about the issues being investigated. These methods helped young participants maintain interest in the process and facilitated comprehension (e.g., feelings identification game when questioned about feelings, i.e., mental health, was essential to guarantee mutual understanding of questions and responses).

The children could engage in qualitative research, but richer data, that is, more comprehensive discussion from participants enabled by more nuanced and thoughtful probing, required advanced capabilities on the part of our interviewers. Children joined in play activities and drawings with relative ease, regardless of drawing ability, it was the interviewers who needed practice to be involved more playfully and to utilize these nonverbal methods to encourage storytelling and contain difficult emotions. More preparation in utilizing art and play as de facto research skills was required for all in the research team. As a result of this, more experienced qualitative interviewers were hired for the main study.

All but one interviewer was surprised at the quality and level of detail and expression the methods revealed in child participants and what they were sharing. Speaking directly from an image produced, created an opportunity to constructively displace loaded emotional content to the page (Lyshak-Stelzer et al., 2007) making it easier to discuss violence. It is vital that interviewers do not interpret what children are sharing in artwork produced or play that gets enacted but rather confirm the meaning directly with the participant to ensure data are accurate and not the projection of the interviewer. Also, for children who may not always feel recognized or respected by adults, having adults validate them through a creation they make, may lead to increased feelings of value and worth as the creation is an extension of themselves (Clacherty & Donald, 2007). Artworks are thus imbued with significance and need to be honored as such. They are also sometimes loaded with content that may, outside of the research encounter, be viewed by others with concern and inadvertently lead to increased safety risks for the child. Asking the child what they would like to do with the artwork once the interview is completed is important to ensure safety.

Considering what the roles and status of children are generally in any given context is crucial. In contexts where children commonly partake in community life through assigned roles and responsibilities but are rarely engaged in decision-making and are seldom asked their opinion, an invitation to participate in research and meaningfully communicate that to children may require extra effort on the part of interviewers. In addition, obedience to and respect for adults are values that are strongly conditioned as a consequence of children rarely being allowed to be forthright with adults (Clacherty & Donald, 2007). Additional time may be required to build rapport with children and highlight that their opinions matter and are valued for these messages to be fully understood and trusted. It was helpful to employ interviewers from the community as they had the language capacity to engage children freely and were more familiar with local games children played. Interviewers need to be deliberate about playing these games and have children address them in a relaxed and appropriate child-friendly way. This translated to sitting on the floor and creating a space with verbal and nonverbal messages of acceptance.

Talking about violence and abuse can be a positive, affirming experience for children. Indeed, research indicates that the same is true for adults and it is not the asking that is concerning but rather how one asks that is most pertinent and “ethically defensible” (Becker-Blease & Freyd, 2006, p. 108). In a meta-analysis regarding adult reactions to trauma research, generally, participation was a positive experience (Jaffe et al., 2015). The silence around abuse experience not only protects abusers and harms victims but it also obstructs scientific discovery (Becker-Blease & Freyd, 2006, p. 225). Discussing violence highlights that a participant’s opinion is significant and respected when often for those who have experienced violence, there may be an internalized lack of importance or worth (Berber Çelik & Odacı, 2020). Also, in typically hierarchical relationships with adults where their perspective is not normally solicited, for children, this experience can be restorative. We saw that this was also restorative for our interviewers (many of whom disclosed in training their child abuse histories) who were more committed, through this process, to take active listening and probing skills into their relationships with their children. But talking about violence can also help to challenge unhelpful notions regarding violence and shift norms or behaviors because it validates experiences and names them as “violence.” When violence is not named by those in authority (usually teachers or parents but in this case interviewers and researchers), then young people may normalize such behaviors (Radford et al., 2017). Helping children name experiences as “violence” enables them to “understand and challenge what has happened to them, to move ‘private’ acts into the public realm, and to alter the boundaries of acceptable and unacceptable behaviour” (Radford et al., 2017, p. 248).

However, the everyday lives of children outside the research context deserve prudent consideration. Research may generate a sense of agency in children that is mismatched with the everyday reality of their marginalization. The realization of this for child participants might lead to consequent disappointment (Parkes, 2008). There is also a chance that increasing the prospect of reflexive agency might motivate a child to challenge existing social relations in ways that increase risks to their safety (Abebe & Bessell, 2014). For example, questioning normalized systems of punishment or violence, or even presenting for dialogue topics previously held private, might be construed as disobedience or provocation, and lead to increased retribution. However, the opposite may also be valid; namely in this research, children knew their parents consented to conversations about violence and that parents themselves were talking to interviewers about this. There may be positive potential to increase dialogue within families regarding violence with potential alternate outcomes.

Sharing personal information and experiences, such as violence, can lead to both an intimate and complex interaction between the child and the interviewer. Interviewers need to negotiate a challenging dynamic of making children feel comfortable and providing support without advertently or inadvertently implying a relationship that cannot be realized. The interviewer will ultimately leave the research site and return to their life, “leaving children behind” and need to be mindful of the implications of the interview encounter (Abebe & Bessell, 2014, p. 130). This is especially true when children have experienced loss, neglect, trauma, or abandonment from adults. Interviewers need to be clear about whether they will return or not and if there will be any contact maintained; this is particularly important in longitudinal research to ensure expectations are managed effectively and with kindness.

Issues of confidentiality needed to be clearly explained to child participants, as well as when confidentiality might be broken in terms of mandatory reporting; but it was also necessary to discuss this with consenting parents/caregivers. Interviews were conducted in the home, an easier context for child interviewees as they were familiar with it, but there was increased interruption in the home, as well as potential for adults to overhear child participant responses or disclosures (homes typically had few rooms), heightening potential for unintended consequences of research as well as safety of the interviewer (SVRI, 2015). Finding a space in the home or outside where parents might be able to “see” child interviews if they had concerns, from a distance but were not able to “hear” them was necessary to ensure confidentiality.

In homes characterized by violence, the violence children and mothers experience are inextricably linked and their relationship is affected by this (Hui & Maddern, 2021; Lapierre et al., 2018). As such, there is a frequent need for multiple forms of support for family members who are victims in tandem. Young children and their families benefitted from our research as any maltreatment or abuse that was reported or suspected received active referrals and management by our social worker, generally with a follow-up that ran over a period of weeks. Ensuring that mandated reports actually served the individual/family identified in the report (i.e., that additional harm was not increased by a referral and that safety was secured or established) was vital to fulfill ethical and statutory obligations and realize positive health outcomes for participants. But, knowing that the disclosure would be attended to, when in many instances this is not the reality (Devries et al., 2015), ensured the interviewer could manage their feelings of helplessness or “rescue” evoked in hearing child abuse or neglect disclosures, and helped maintain the mental health needs of interviewers (Jewkes et al., 2000).

Treatments and policy for children are often based on the outcomes of studies on adults and scientifically evaluated treatments and policy are less available to children due to barriers to their inclusion in research studies. Children and young people can offer perspectives that strengthen adult capacity to understand their lives and these insights are vital to accurately identify the need and development of service innovation and progressive policies (Grace et al., 2019). Initiating violence treatment and prevention interventions and policies in early childhood may be critical factors in breaking the widespread cycles of violence, especially in LMICs.

Our experience shows that meeting high ethical standards of research with young children is possible when multiple participants are recruited from within families experiencing violence simultaneously. This does, however, place major demands on interviewers and their managers, who require capacities and attitudes compatible with these demands. This has cost implications and funders need to be alert to this reality.

Limitations

A key limitation to this research is we did not ask the children more fully about their experience of participating in it; something we will be addressing more meaningfully in the main study. We also did not conduct direct observation so our presentation of the findings relies exclusively on the interviewers’ subjective interpretation of the interview process and the children’s reactions. However, the Supplemental Appendix may offer some insight into the process and outcomes with examples of interviews. Young children respond well to certain types of question frameworks, such as “yes” and “no,” but particularly the younger participants had challenges giving more elaborate explanations to questions of participation in this research. However, we believe that this limitation was mediated by the use of arts- and play-based approaches that seemed to yield more detailed storytelling, expression, and communication of experience. Thus, we did not rely solely on verbal methods to understand the children’s experience of the interview. This was a rural sample and perhaps there is more variability or key differences in an urban sample. Interviewers had limited research interviewing skills, likely constraining potential findings. Child participants, as well as FGD participants, might have been biased by social desirability.

Recommendations

Violence is a particularly diffuse experience and resultantly its definition is complex to operationalize. Beliefs about what is acceptable and unacceptable in terms of behaviors and what comprises harm are culturally influenced and continually under review as values and social norms evolve (Fraga, 2016). Consequently, researchers need to explicitly identify and convey the values on which their research is based; this was relevant in our study to principles around both violence and child participation in research. In calling for recognition of values, we as researchers are also promoting particular values ourselves and these need to be transparent and openly stated (Abebe & Bessell, 2014). These messages bear repetition to ensure they are understood over time.

Interviewers must engage in playful behavior, meet children at their level, and ensure the research is fun and engaging. This expands Mullender et al.’s mnemonic of three Cs (consent, confidentiality, and child protection), and three Ds (danger, distress, and disclosure) to include children’s position and agency, expressed by the three Es (enjoyment, empowerment, and emancipation) (Houghton, 2015; Mullender, 2002). It is helpful for interviewers to stress how interested they are in hearing what the children have to say, how important this information is, and that the child is appreciated for their original contribution (Ponizovsky-Bergelson et al., 2019). These guiding principles were essential in our study and were not obviously understood or internalized so required practice to enhance positive research outcomes.

Interviewers need to recognize that “all behaviour is communication” especially with regard to young children and feelings of emotional overwhelm that may show up in children’s bodies with “difficult to manage or oppositional behaviours” or where children may become quiet, looking distracted, or focusing on something else in the space. This requires extra patience on the part of the interviewer and intentional management of the behavior so that the participant can regulate themselves, feel safe, and get back to the interview or terminate it, whichever is in the child’s best interest.

When facilitating drawings with children, it is important to allow them enough time to complete the task comfortably (this is true too for completing the questionnaires). Because adults may be impatient or mistakenly think that the time taken by children to think is silence, refusal or not knowing the answer, this requirement needs to be intentionally addressed in training. It is helpful to state upfront that there are no right or wrong answers, and no right or wrong ways to draw, offering encouragement and enthusiasm for task completion throughout the interview.

Disclosure is related to the skills and sensitivity of the interviewer and resultantly, the team should be purposively selected. When interviewers ask child participants to share stories around violence that have trauma content, there is an opportunity to listen in a sympathetic, supportive, and non-judgmental way. Although interviewers should not take on a role as counselors, they should be open to assist if asked and know when to refer child participants for additional support. Recruiting for “sensitivity” in fieldworkers is important. There is a need to recognize the emotional effect that listening to these kinds of stories, particularly with young children, can have on interviewers (SVRI, 2015). Ongoing management of how to deal with emotional costs and ethically challenging dilemmas associated with violence research needs to be discussed openly within teams to ensure interviewers do not become overburdened or numb to the stories and that their health and the quality of their interviews are not compromised. In our study, there was open discussion in team meetings about this and regular check-ins with staff by the project manager and investigative team to ensure fieldworker mental health was recognized and supported.

Depending on the length of time of the study, it is helpful to have a social worker as part of the team if there are mandated reporting requirements that need to be actioned, with active referral and ongoing follow-up to ensure service delivery is provided and legal obligations met. It is not typical to hire social workers into research teams, but in this study, it was not only helpful but vital to ensuring participant needs were met ethically. Contextual realities are also implicated in ethical obligations for the research team and in areas where unemployment is rife, and food insecurity is high, the team may need to approve supplementary support that may be indicated and budget appropriately (Devries et al., 2015).

Ensuring two interviewers are present for each home visit increases interviewer safety but also enables child participants to increase confidentiality in home settings. While one interviewer interviews the child participant, the other can either interview a caregiver or keep them company. This recommendation stems from other research in South Africa, but in our study with young children and within a multigenerational context, it was necessary (Jewkes et al., 2000; WHO, 2001).

In conclusion, all children have the right to engage in research. Research on violence within families presents many ethical and practical challenges but our pilot study proved that children as young as 4 years old can engage meaningfully in it and local interviewers, with little prior experience, can be trained and mentored in engaging child participants effectively. We hope our findings may add to the field of violence research in LMICs where research involving children is sparse and where concern for welfare is amplified.

Supplemental Material

sj-docx-1-jiv-10.1177_08862605241230088 – Supplemental material for “Children Are Like Vuvuzelas Always Ready to Blow”: Exploring How to Engage Young Children in Violence Research

Supplemental material, sj-docx-1-jiv-10.1177_08862605241230088 for “Children Are Like Vuvuzelas Always Ready to Blow”: Exploring How to Engage Young Children in Violence Research by Nataly Woollett, Nicola Christofides, Hannabeth Franchino-Olsen, Mpho Silima, Ansie Fouche and Franziska Meinck in Journal of Interpersonal Violence