Fieldwork on Intimate Partner Violence under Community Opposition
A university team plans a household survey on intimate partner violence among 400 married women in a rural block. The questionnaire includes a module on sexual violence. Interviewers are local women with two days' training; one is related to a respondent's husband. The panchayat has announced that the survey will break families, and a local youth group has threatened to stop fieldwork. The ethics committee asks how the team will handle disclosure of violence, mandatory reporting and referral; the nearest One Stop Centre is 45 km away, and there is no counsellor on the team. The budget covers ten days of fieldwork. Many women are home alone only in the afternoon. The consent form is on a tablet, but about a third of women have low literacy.
1. Design a safety and ethics protocol that meets feminist research standards without making the survey impossible to field.
Read the worked answer
Use the WHO ethical and safety recommendations for research on violence against women. Interview in a private space without the husband or other relatives. If privacy is impossible, skip the violence module and record why. Obtain informed consent in the local language, making clear that participation is voluntary and can stop at any time. Train interviewers on distress, confidentiality and non-judgmental questioning.
Have a referral list with the One Stop Centre, helpline, local NGO, police and health centre, and a protocol for immediate danger. Explain the limits of confidentiality, especially where a child is at risk or a weapon is involved. Do not make mandatory reporting a precondition for help. Secure data: no names on tablets, encryption, and a separate ID key. Protect interviewers through pairs, check-ins and no evening travel. Form a community advisory group that includes local women. Because the One Stop Centre is 45 km away, include transport and phone follow-up in the plan.
2. How should the team respond to the panchayat's opposition while preserving scientific validity?
Read the worked answer
Do not seek the panchayat's permission as if it were a gatekeeper for women's participation; that would let male institutions control the sample. But do engage. Explain the study's purpose and confidentiality, and make clear that no individual results will be shared. Meet women's self-help groups separately and offer a community feedback session on aggregate findings.
If opposition threatens safety, change venues to an anganwadi, health sub-centre or NGO office, use female enumerators from outside the immediate village, and avoid advertising. Document the opposition as part of the research context. Validity depends on not sampling only women whose husbands allow it. If access is blocked, report the bias and do not generalise. The team should have a safety plan and a clear decision rule to pause fieldwork.
3. The team finds a 4% prevalence of sexual violence but suspects underreporting. What methodological steps can improve disclosure, and what claims should they avoid?
Read the worked answer
Underreporting is likely. Improve disclosure through careful wording that asks about specific behaviours rather than labels, build rapport, ensure privacy, and use self-administered audio computer-assisted methods where literacy allows. Train interviewers to avoid judgement and to ask about sexual violence later in the interview. Check whether the 4% is driven by one village or by interviewer effects. Consider a list experiment or indirect questioning in a subsample.
Triangulate with qualitative interviews and service data, but do not treat service data as prevalence. Report the estimate as a minimum and give a range. Avoid claiming a true prevalence or ranking villages. The claim should be about patterns and associations, not a precise figure. Do not publish village-level identifiers. State the limitations clearly in the final report.