Data sources and collection

Quantitative data on violence against women and girls 

• Reliable statistics on the prevalence, causes, and consequences of all types of violence against women and girls are essential to developing effective legislation, to advocate for the legislation, and to develop strategies and protocols for the implementation of the legislation. See: Researching Violence Against Women:  A Practical Guide for Researchers and Activists.

For example, UNDP Albania, UNICEF and UNDP Small Arms Project supported the Albanian Institute of Statistics in conducting a national domestic violence survey in Albania. The survey report, Domestic Violence in Albania: A national population-based survey was published in March of 2009. The goal of this significant project was to establish national baseline data on domestic violence in Albania. It is the first of its kind ever done in Albania.

• Statistical data should be obtained at regular intervals.

• Quantitative data can also be obtained by measuring indicators of government response and program performance. This research should include measures on numbers of prosecutions, convictions, denials of prosecutions, orders for protection issued, and re-offense data.

• Quantitative data can also be obtained by coding qualitative data into measurable units. This can help to describe and summarize large qualitative data such as hundreds of interviews. See sections on Qualitative data and Analyzing data from interviews.

• Data should be disaggregated by sex, race, age, ethnicity, relationship between perpetrator and victim, and other relevant characteristics. Government offices should be required to compile and analyze this data. See: UN Handbook, 3.3.2.

For example, the county of Kazakhstan has created an automated databank in the Department of Public Safety of the Ministry of Internal Affairs, which allows the government to obtain data on the age of the victim, the consequences of the violence, and the criminal disposition of the case, among other factors.

• Governments should sponsor or create a baseline study on the relevant issue of violence against women and girls as a reference point for future studies.

For example, the Baseline Survey on Domestic Violence and HIV (2008) conducted by the Centre for Media Studies and Breakthrough of India anchored a three-year campaign to combat domestic violence and discrimination against women with HIV/AIDS.  The study created benchmarks on existing knowledge, attitude and practice about domestic violence, HIV/AIDS, and the Protection of Women from Domestic Violence Act (2005) of India, and on attitudes on inter-spousal communication, condom use and negotiation, and sources of information about these issues.

The UNFPA and the National Statistical Service of Armenia implemented a quantitative survey to collect data on gender-based violence in 2009-2010, in response to the need for official data on violence against women and girls, so that the importance of addressing the problem could be substantiated. They sought to create an official baseline for policy makers in Armenia to utilize in the formation of policies and programmes to reduce gender-based violence. Survey administrators collected representative information at the Marz, or regional level in Armenia. (There are 11 marzes in Armenia.) Using the database of addresses from the 2001 population census, field workers visited nearly 5000 households and administered questionnaires on gender-based violence. Data was coded and entered into a computerized database. A working group, consisting of Ministry officials, police, and members of civil society, led by a sociologist and gender expert, was created to analyze the data. The results are available in English. 

Qualitative data on violence against women and girls

Qualitative data is data that goes beyond statistics. It is data from interviews, descriptions, case studies, and observations. The United Nations Secretary-General’s in-depth study on all forms of violence against women (2006) stated “Qualitative methods are necessary to complement quantitative surveys, for example to understand the complexities and nuances of experiences from the respondent’s point of view…qualitative research findings are useful in assessing women’s needs and constraints, community needs, designing prevention campaigns, planning and evaluating interventions and engaging community actors via participatory research.” Sec. 217

Qualitative data can be obtained by a number of methods:

• Interviews with survivors, advocates, attorneys, police, prosecutors, judges, probation officers, health professionals, and social service providers.

• Focus groups, such as groups of service providers. For more on focus groups, see: World Health Organization and Program for Appropriate Technology in Health (PATH), Researching Violence Against Women:  A Practical Guide for Researchers and Activists (2005), Chapter Nine.

For example, the Ukrainian NGO Progressive Women held a focus group with survivors of domestic violence, asking them “In your view, what would be the perfect resolution of domestic violence by the state?” The NGO reported that victims had creative ideas: the emergency number should be 2 or 3 digits that are closely spaced on a telephone so that the buttons are easy to push in a stressful situation; a male and female police officer team should arrive within 15 minutes; victims should be questioned in a room away from the perpetrator; specialized courts should consider the case; penalties should include community service and a stay in a special “Center for Rehabilitation of the Offenders;” and social services should offer victim support for many months.  Importantly, victims wanted domestic violence to be classified as a criminal offense, so that responsibility for collecting evidence would be transferred from the woman to the state.

• Participant observation, wherein a researcher sits in a group setting in order to gain firsthand knowledge, such as observing a victim support group. This is a good way to discover differences between what should happen and what actually does happen in implementation of the law.

For example, in Domestic Violence in Brazil: Examining Obstacles and Approaches to Promote Legislative Reform (Jodie, Roure, 2010), the researcher observing a focus group noted that victims felt a sense of hopelessness and lack of judicial protection because of the government’s ineffective actions in domestic violence cases. One victim asked “how she would explain to her children that it was morally wrongful to batter when there was no judicial authority to support this message.” p. 88

For example, when activists in Ukraine investigated the question of why women did not report violence to the police, they accompanied police patrols on domestic violence cases, interviewed victims, and observed trials. They observed first hand the delay that victims experience when they call the police, and that the sanction most often given was a fine, which, since it came from the family budget, worsened the financial situation of the victim. They observed that the perpetrator, detained for only 12 hours, continued to live with the victim in most cases. They observed that regular physicians did not want to do the paperwork to document injuries, and that forensic doctors may examine a women only by police request, and the procedure to obtain this request is very difficult.  From: Ukrainian NGO Progressive Women.

• Surveys
  For an example of a qualitative survey on domestic violence, see: Security Begins at Home: Research to Inform the First National Strategy and Action Plan against Domestic Violence in Kosovo (Nicole Farnsworth and Ariana Qosaj-Mustafa for the Kosova Women’s Network, 2008), Appendix 2.

• A search of court records
  See section on Monitoring court records, below.

• Observation of court processes
  See section on Court monitoring, below.

• Media reports
  Gender-Based Violence In Tanzania: An Assessment of Policies, Services, and Promising Interventions (Myra Betron, 2008) is an example of a qualitative monitoring report. Methods used included key informant interviews and focus group discussions. In this report, monitors found that women and girls are frequently blamed for causing or provoking gender-based violence, and that victims rarely report the violence to law enforcement, seek treatment, or visit support services. The assessment also describes gaps in Tanzanian legislation and in capacity of service providers.  Promising interventions are also noted in the report.

Sources of data on violence against women

Data may be obtained from surveys, administrative statistics, or criminal statistics:

• Data may be obtained by a dedicated survey which asks questions only on violence against women. Dedicated surveys will result in more extensive data on violence against women but are more expensive to administer and costly to repeat regularly.

• Data may also be obtained by a survey on a broader topic, such as health, which includes a section on violence against women. For example, one survey on crime in England incorporated questions on violence in the home. See: Safety and Security: A Proposal for Internationally Comparable Indicators of Violence (CRISE, 2008). The National Family Health Survey Report of India (2005) also incorporated questions on the percentage of women who have experience domestic violence in intimate partner relationships.

• National statistics offices are important sources of data on violence against women. These offices may administer surveys and they may collect data from the criminal justice system.

• Data can also be obtained from relevant administrative sources. Administrative sources include health providers, the criminal justice system, social service providers, public housing providers and others.  However, these administrative sources may pose challenges to developing a comprehensive picture for several reasons, including:
  
  1. They only capture data from those who report the abuse they have experienced, and the majority of women and girls do not report the abuse therefore data that exists is a gross underestimation of reality.
  2. The data is not disaggregated by gender, age, or relationship.
  3. Streamlining the data collection across the nation is challenging as is ensuring consistency in the execution of data collection.

Data should be collected according to United Nations Statistics Division, Fundamental Principles of Official Statistics, available in Arabic, Chinese, English, French, Russian, and Spanish.

Combining quantitative and qualitative monitoring

When monitoring legislation on violence against women and girls, it is most effective to utilize a combination of qualitative and quantitative monitoring. Quantitative data on the number of complaints and cases of domestic violence which reach the court system is more meaningful when combined with qualitative data such as interviews with advocates or survivors who have had experience with barriers to court access which survivors experience. Utilizing both methods of research allows a researcher to identify inconsistencies and truths which may not be immediately evident when only one type of data is employed. 

This type of monitoring requires a “situation analysis” which describes multiple factors around the issue of violence, including: specifics of the law; interviewees’ reports of experiences of victims of violence; and the government response to the violence. The report should evaluate the adequacy of the government response in attaining victim safety and offender accountability, and make recommendations to all stakeholders.

See: World Health Organization and Program for Appropriate Technology in Health (PATH), Researching Violence Against Women:  A Practical Guide for Researchers and Activists (2005), Chapter Three.

For example, the Baseline Survey on Sexual and Gender Based Violence in Rwanda (UNIFEM, 2008) combined quantitative data on sexual and physical violence with qualitative data that examined the effects on survivors and evaluated stakeholder actions to address and prevent gender-based violence.