The Importance of Contextual Factors in Carrying Out Childhood Violence Surveys: a Case Study from Indonesia.

Nationally representative research into violence against children is necessary to understand the scale and complexity of such violence and to evaluate prevention efforts. To date, however, most countries do not have adequate data. In 2013, the government of Indonesia conducted a national Violence Against Children Survey (VACS). This was a cross-sectional household survey of male and female 13-to-24- year-olds designed to estimate physical, emotional, and sexual violence prevalence. The target was to interview at least 2580 individuals; but response rates were much lower than anticipated (females = 66.6%; males = 56.1%). Insufficient data was available across several variables and there were unexpected anomalies in obtained data. We conducted a retrospective analysis of the survey to understand impediments and to advise future national efforts in Indonesia and other low-to-middle-income contexts. Survey managers and implementers ( n  = 22) were interviewed online and in person. We also carried out secondary analysis of the child survey data to identify factors possibly associated with (non-)response and assessed field notes from interviewers. Culturally inappropriate timing of data-collection (during Ramadan) may have had a negative impact on household responsiveness and the availability of children at home. Face-toface interviews in households were considered to impede participation and disclosure. Survey field staff and managers expressed the need for deeper training and a more comprehensive pilot. Recommendations to improve privacy and anonymity include the use of self-administered questionnaires and school-based rather than at-home surveys. These and other findings from this case study may be useful in planning future surveys in Indonesia and similar social and cultural contexts.