Study title

Cambodia

Household

Individual

Quantitative household survey: The quantitative household survey was conducted in four districts (two rural and two urban) in Pursat and Siem Reap Provinces. Pursat was purposively chosen because project partner Gender and Development/Cambodia (GAD/C) operates there and it had high occurrence levels of domestic violence. Siem Reap was selected because the project’s Principal Investigator (Dr Katherine Brickell) had conducted research on domestic violence in the province previously (with the opportunity for longitudinal data noted). A total of 1,275 respondents were surveyed. The design used for selecting respondents was based on a multistage stratified sampling technique, with the probability proportional to size (PPS) method for selecting villages and the systematic sampling technique for selecting households. Once the districts were identified and selected, a list of communes was generated. Two communes were then randomly selected from each district. Next, a list of villages with the number of households taken from the 2008 General Population Census data was compiled separately for each selected commune. PPS was used to select five villages for each commune, resulting in a total of 40 villages selected for the study. The next step was to select households and respondents for interview. Information about the number of villages was taken from 2008 Census data and was used as the basis to calculate the sampling interval for selecting households. A random start was used from the village boundary. After selecting a household, two adults were targeted asked for interview. While primary interviewees included the household head and his/her spouse, in circumstances when one primary respondent was absent or in non-spousal households, another adult living in the selected household could be selected who was a different sex (it did not matter if they were of the same generation). A bespoke household survey was developed for the study and tested in a pilot study prior to the substantive research. Interviews were held face-to-face and conducted by 8 trained interviewers and one field supervisor. Stata has been used for data analysis. Participatory video (PV) research: Four PV workshops were held in total (divided equally between the two provinces in 4 selected districts). The workshops were 3-days long each in total and were held with lay men and women to understand community-level perspectives on domestic violence (ordinarily considered a 'private' issue in Cambodian society).Lay and stakeholder interviews: Divided equally across the same districts as the PV research, qualitative data was collected from forty interviews with female domestic violence victims; forty with male and female householders; and a final forty with stakeholders who have an occupational investment in domestic violence alleviation (this included legal professionals, NGO and health workers, police officers, religious figures, and provincial, district, commune and village-level authorities). The research was conducted by two research assistants (one female and one male given ethical considerations). They were provided with information on NGOs and other services which were available to domestic violence victims should they request this information. Phnom Penh-based institutional stakeholders: At the end of the research period, ten interviews were held with high-level NGO advocates and government representatives.