The first time we met Yati (pseudonym) was during a focus group discussion (FGD). Through a local woman leader, we invited a number of women to discuss their knowledge of domestic violence and domestic violence reporting services. Yati was invited as a representative of the women in her community, not as a victim or survivor of domestic violence.

Naturally, we did not delve deeper into those personal experiences of domestic violence during the public FGD. We closed by thanking the participants, and asked permission to contact them in future if needed.

After the FGD, we visited Yati’s house and she kindly welcomed us. Unfortunately, her husband was interested in joining our conversation, so we were not free to talk about domestic violence. So we said then that we only stopped by to say hello and we invited her to have another discussion some other time. Yati understood our ‘code’, and she offered an alternative time and place.

In private Yati spoke openly of her experience and reporting on domestic violence. Yati trusts that all information she provided will only be used for research purposes and her identity will be kept anonymous.

The above is a snippet of our experience in conducting a longitudinal study on domestic violence reporting as part of a larger longitidunal study on poor women’s access to public services.[1] Although members of a community might know domestic violence exists, there are many reasons they might not be free to speak about it openly. At a social level, domestic violence is often considered a personal matter, only appropriate for private discussion within the family. If it is discussed outside the family, taboo can prevent discussion with people outside the community, such as researchers. For certain people, it can also bring public shame and harm their family’s good standing in the community. Domestic violence can stir uncomfortable emotions or dredge up traumatic memories, both in study participants and researchers. It is important for researchers to consider these limitations and sensitivities in their study.

Building trust of study participants through a multi-step approach

It is not easy to extract information about sensitive issues such as domestic violence. The main lesson learned from this study is the importance of building trust between participants and researchers to allow participants to be more open in sharing what they know and have experienced, and their hopes and perceptions about domestic violence. When trust has not been established, participants would tend to give normative answers that may not reflect the real situation, both of any domestic violence they have personally experienced and of what is happening in their community.

Building trust between researchers and participants is not an intantaneous process. Throughout the 3 data collection temporal points (2014, 2017, and 2020), we attempted to develop various approaches to build trust with the participants. We approached them in 3 ways.

First, researchers prioritised sensitivity when interacting with participants. We realised that sensitivity in studying domestic violence is not something that just grows easily. Before collecting data, researchers underwent training to form an understanding of domestic violence issues, interview techniques and FGD facilitation techniques. Training materials were taken from reading references and stories shared by practitioners from the government and non-governmental organisations (NGOs) that work in victim rehabilitation, handling cases of domestic violence and anti-domestic-violence programs.

Our evaluation shows the importance of opening surveys, interviews and FGDs with general questions, before honing in on discussions of domestic violence. When exploring domestic violence issues, researchers need to remain sensitive to participants’ responses, such as their intonation, volume, facial expressions and body language. If researchers perceive or observe changes, they should consider changing their approach and manner of asking questions and in more extreme cases delaying or even stopping the inquiry outright.

Studying domestic violence issues demands great patience from the researcher. They must be able to listen without pushing participants to share or interrupting their responses. The data collection process is not a unilateral interaction of researcher asks, participant answers; participants need to get feedback from the process. For example, at the end of a household survey, the interviewer can tell the participant what actions they can take if they become victims or learn about incidents of domestic violence in their community.

Second, involving local leaders. When researchers approach participants in a way that is acceptable to the local community, or enter a community with the help of people who are known to the participants (such as village officials and women leaders or village NGO cadres), initial trust in researchers will be established. From our experience, the figures involved should be those who are used to helping the community, or have helped participants when they experienced or reported domestic violence. This can pave the way to good communication and build participants’ trust.

Third, building a safe and comfortable space and atmosphere for participants to share stories. Steps taken towards building a comfortable space and atmosphere include obtaining participant consent, guaranteeing anonymity and confidentiality of all information and ensuring the legality of all research activities. All are essential when collecting data on sensitive topics. The researcher submits an informed consent form to the participants containing information about the study, the risks and benefits of participating, as well as guarantees for the confidentiality of participants’ identity and personal information. The researchers should only start the inquiry – including voice recording and note-taking – if the participants have given their consent. Photography of any kind is to be avoided. The data and information confidentiality guarantee also applies to all forms of publications produced, even after the study ends.

We recognise the importance of protecting research participants, especially on sensitive topics such as domestic violence. All guidelines were submitted to the ethics committee of the Atma Jaya University Research and Community Service Institute (LPPM) to ensure that there are no questions that might threaten participants. In terms of the legality of the activity, we ensure that permits from national to village level governments have been secured before commencing data collection. This is important not only to secure access and freedoms for researchers formally, but also to provide participants with a sense of security, comfort and assurance that our activities are legal and that we are known to the local government.

Prior to data collection, we first tested the entire guide. From the results of the trial, we realised that information about participants’ personal experiences as victims, survivors or reporters of domestic violence should be collected through private and anonymous interviews. FGDs should be conducted to explore the condition of the community; not to explore someone’s personal experience. The FGD atmosphere must be made as comfortable as possible, without cornering or embarrassing anyone.

The ‘gossiping’ method is effective for starting an exploration domestic violence during discussions. When a participant mentions a domestic violence case that they know about and then discusses it, this will gradually prompt other participants to share incidents that happened to them. In this situation, the researcher needs to be an active listener, not patronising or judgmental, and try to understand the victim’s situation. In the household survey, participants’ answers were kept confidential in an envelope that will only be opened by SMERU researchers. Even the enumerators do not know participants’ answers.

Design of longitudinal study allows for repeated interactions between participants and researchers

The longitudinal study design strenghtens the quality of research into the issue of domestic violence. Repeated interaction process at several temporal points of data collection fostered participants’ trust in the researchers, the same research location was used for the duration of the study, and in survey activities, researchers always involved the same households as participants. Interview and FGD activities also involved the same participants. In addition, local figures and researchers were involved from the beginning of the study. Trust was key to participants’ openness to sharing stories about their experiences and knowledge, so researchers could obtain more comprehensive information to produce quality policy recommendations.

Longitudinal studies also allow researchers to evaluate and update guidelines based on prior performance. For example, learning from data collection during the baseline study (2014), it became apparent that focusing interviews solely on survivors provided less information than expected. We improved on this for the midline (2017) and endline (2020) data collection. In another example, when there was a local researcher who was not sensitive about domestic violence for their support of polygamy, that became our grounds for not involving them again in subsequent data collection.

Combining quantitative and qualitative methods to obtain more comprehensive information

We realise that research data could be made more complete and capture the needed information by using more than one method. In the baseline study, only qualitative data was collected, but we learned that information from interviews alone was not enough, especially with limited time for data collection. We also examined other studies and found that surveys and FGDs were useful tools for collecting data on sensitive issues such as domestic violence. Thus, during the midline (2017) and endline (2020) studies, we combined quantitative and qualitative methods by conducting household surveys, interviews and FGDs.

Researcher evaluation showed that the use of varied collection methods will gather complementary data. Household surveys will provide a simple number on the prevalence of domestic violence and domestic violence reporting behaviour. Meanwhile, interviews and FGDs complement this with data on reporting flows and influencing factors. Data collection carried out at different levels – individual, household, community, village and district – can also provide a more comprehensive picture about the domestic violence situation at study locations.

In the end, data collection on sensitive issues such as domestic violence – at any level and using any method – demands the trust of study participants. Their trust in researchers is the key to information opennes to obtain comprehensive data. However, the process is not instantaneous, and requires an understanding of issues, creativity in data collection, and certainly the willingness of researchers to continue to learn to empathise with study participants. Although it is not easy, building trust in domestic violence studies can yield more detailed and useful data.