Gender-based violence in conflict-affected settings: overview of a multi-country research project

Conducting GBV prevalence research in conflict-affected settings presents a host of scientific, ethical, security and methodological challenges.

Gender-based violence (GBV) describes any harm perpetrated against a person's will that is rooted in power inequities informed by gender roles. GBV encompasses physical, sexual and psychological violence, threats of violence, coercion or arbitrary deprivation of liberty. Though GBV may take many forms, it almost invariably disproportionately affects women and children. Humanitarian agencies have become increasingly concerned about the extent and effects of GBV in refugee, internally displaced and post-conflict settings. GBV is belatedly being recognised as an affront to public health and human rights principles and a major impediment to refugee/IDP reintegration.

However, increased awareness of the problem is yet to be matched by either consistent data collection or dissemination of best practices for addressing GBV. To meet these gaps and to improve international capacity, the Reproductive Health Response in Conflict (RHRC) Consortium launched the Gender-Based Violence Initiative in 2000. Major outcomes have included the first comprehensive global overview of GBV issues affecting displaced people and a manual to improve programme design, monitoring and evaluation.

The GBV Tools Manual contains a working draft of a standardised population-based survey designed to measure multiple forms of GBV in conflict-affected settings. Field-tested in East Timor and Kosovo, the questionnaire was subsequently used to conduct a national survey in Rwanda and a survey among IDPs in Cartagena, Colombia. The four studies have sought both to generate reliable prevalence data for locally-based programmes to use in GBV-related planning and advocacy activities and to build local capacity to conduct population-based research. Local partners are being provided with key tools, such as a validated questionnaire in the local language, necessary to design follow-on research projects. Data are being generated to enable the first multi-country comparisons of rates of GBV in conflict-affected settings.


The multi-agency research team, including researchers from the University of Arizona, the US Centers for Disease Control and Prevention (CDC), and the RHRC, prepared the questionnaire to facilitate comparability with existing questionnaires by including, wherever possible, previously tested questions or response options and adapting them to conflict settings. The questionnaire was designed to be locally customised without undermining the reliability and comparability of standard measurement. It is divided into sections that focus on different time periods and types of violence. Each can be removed in its entirety according to the objectives of the investigators. Prior to field use in each country, the questionnaire was reviewed by a team of local partners who made appropriate revisions. The questionnaire was then translated into the local language, back-translated into English and checked for accuracy. Further revisions were made following pilot testing.

Choice of country sites was determined on the basis of: 1) lack of pre-existing data on the nature and scope of conflict-induced GBV; 2) existence of local programmes with field research capacity and ability to provide referral services to participants; 3) the need to ensure representative global coverage; 4) potential for local follow-up advocacy; and 5) considerations of accessibility and security.

In each country women of reproductive age were randomly selected to participate in the surveys. In East Timor, Kosovo and Rwanda, population lists were obtained from local officials and population-proportional samples were selected. In Colombia, where no population lists were available, sample selection was based on a mapping of households. For all countries, only one woman from each selected household was asked to participate.

For the field tests in East Timor and Kosovo, as well as for the national research in Rwanda, a detailed research protocol was submitted for review at CDC. A summary of the Rwanda protocol was also submitted to national government partners. In Colombia a local ethics committee reviewed and approved the protocol and questionnaire. The translated questionnaire was presented to team members for feedback. Interviewers practised administering the questionnaire among themselves and then conducted pilot tests among a sample of women. The pilot test gave the supervisors an opportunity to assess the skills of the interviewers and to make a final selection of the interview teams. Based on the pilot tests, final revisions were made to the questionnaire.

Field-based partners, including international, national and local NGOs, were actively involved in all aspects of research planning and delivery. Efforts were made to recruit all-female research teams from local women's organisations and to include representatives from target research populations. Research teams received two weeks training. Rwandan and Colombian teams included 'psychosocial assistants' to address any issues that might arise for interviewers or participants during the interview process. Data collection forms were stored and locked each day and any potential identifiers were removed from research materials to preserve participant anonymity. All team members had to sign a confidentiality agreement.

Because of the sensitive nature of the questions and the difficulty in obtaining privacy at the participant's home, women who were willing to participate were interviewed at a central location outside the participant's home. Informed verbal consent was obtained from each woman. Where appropriate local health and psychosocial services existed, participants were informed that referrals were available and a list of organisations offering support services to survivors was provided on request. In Rwanda, participants were advised to access local women's representatives who had been apprised of the research and had agreed to provide follow-up support if necessary.

In East Timor and Kosovo the research teams concluded that the number of women agreeing to participate in the research was inversely related to the degree of visibility of the research project. The higher the visibility of the research, the less likely women were to consent to participation or to show up for the interview. The research design was adjusted with positive outcomes in Rwanda so that the researchers were only working in a village for an average of one day. In Colombia, as an additional security precaution, the interviews were conducted outside the barrios; however, this presented its own challenges as fewer women were willing to travel the distance required to be interviewed.

Field data from East Timor and Kosovo was entered and analysed at the CDC while partner organisations in Rwanda and Colombia are analysing their data in-country. Research findings from East Timor - the only country where data has been finalised - were disseminated by participating international and local agencies through focus groups and the national media.

Overview of findings: East Timor

A quarter of the 288 women who participated in the East Timor pilot study reported exposure to psychological and physical violence perpetrated by a non-family member during the crisis-related violence that followed East Timor's 1999 vote for independence from Indonesia. Most GBV victims reported being threatened with a weapon and subjected to abusive sexual comments. In over two-thirds of cases, women were threatened with death by members of local militias or the Indonesian military or police.

Levels of reported non-family member violence were significantly lower for the post-crisis period, with a 75.8% decrease in physical violence and a 57.1% decrease in sexual violence, though types of violence most commonly reported stayed relatively constant. Displacement to a camp in West Timor was significantly associated with reports of post-crisis sexual violence. After the crisis had passed GBV perpetrators within East Timor were primarily identified as neighbours and other community members.

Levels of intimate partner violence were investigated for two periods: the year before the crisis and the 12 months prior to administering the interview. 46.8% of all women in relationships reported some form of intimidation and control, verbal abuse, physical assault or sexual coercion by their partner in the year before the crisis and 43.2% in the past year. Among women in relationships, 23.8% reported physical assault in the year before the crisis and 24.8% in the past year. Of the women who had ever experienced domestic violence, 41.5% sustained physical injuries but only a third of those women sought medical treatment for their injuries.

Findings on help-seeking behaviour suggest that East Timorese women most often seek assistance from family members. For crisis and post-crisis outsider violence respectively, 6.9% and 13.3% of women who experienced violence reported it to the authorities. Of those who did not tell anyone about their experience, 38.7% (during crisis) and 50% (post-crisis) did not tell because they believed nothing could be done. Domestic violence survivors were even less likely to seek assistance than survivors of violence perpetrated by someone outside the family.

In East Timor the research methodology has informed ongoing national GBV research and pilot test findings have fed into parliamentary discussions on how to address GBV. It is hoped that similar positive outcomes will follow from the release of the data in Kosovo, Rwanda, and Colombia.

The way forward

This multi-agency, innovatory and global collaboration has demonstrated that:

  • With sufficient planning, training of researchers and time for rigorous pre-testing it is feasible to carry out GBV prevalence research in conflict-affected settings.
  • It is possible to design a survey questionnaire and conduct population-based research using methodologies meeting international standards for reliable data collection while supporting local partnerships and ensuring local ownership of knowledge generated.
  • Local researchers lose their initial hesitation about asking prying questions in settings where GBV is perceived as a private issue: post-research debriefings were universally positive, with many researchers feeling that the interview provided an unprecedented opportunity for participating GBV victims to receive validation and support.
  • Effective risk reduction strategies can be developed in collaboration with local partners - in none of the countries did researchers face any security incidents.


Jeanne Ward is the GBV Research Officer at the International Rescue Committee ( To learn more about the RHRC GBV Initiative, see or contact the author. Email:


Opinions in FMR do not necessarily reflect the views of the Editors, the Refugee Studies Centre or the University of Oxford.
FMR is an Open Access publication. Users are free to read, download, copy, distribute, print or link to the full texts of articles published in FMR and on the FMR website, as long as the use is for non-commercial purposes and the author and FMR are attributed. Unless otherwise indicated, all articles published in FMR in print and online, and FMR itself, are licensed under a Creative Commons Attribution-NonCommercial-NoDerivs (CC BY-NC-ND) licence. Details at