The first step is to identify which types of data require improvement and how. For example, prevalence data, which provides insights into the scale of violence within a population, is distinct from administrative data, typically derived from service providers or law enforcement, which only provides details on reported incidents and responses. Each of the following data types serves a unique purpose in addressing VAWG and requires tailored strategies for improvement:

• Generating accurate prevalence data requires comprehensive and statistically representative population surveys.
• Improving the accuracy of administrative data may require improving national or subnational data coordination mechanisms.
• Research studies, another key data type, delve deeper into the underlying causes, including how different groups of women and girls are impacted. It is key to ensure that these studies have rigorous academic methodologies - whether quantitative, qualitative or mixed methods - to ensure validity and reliability.

Adequate budgeting for improving VAWG data is essential for long-term viability and impact. Resources need to be allocated for design, testing, data collection, and dissemination, as well as the ongoing maintenance of data collection systems, ensuring they remain up-to-date and effective over time. This includes costs associated with training personnel in data collection and analysis, the technological infrastructure required for storing and processing data, and the tools necessary for effective data gathering, such as surveys and reporting mechanisms. In addition, resources are needed for the effective dissemination of data, involving data visualisation and sharing data with relevant stakeholders to ensure the data is used to inform policymaking and programme design.

A key element of improving VAWG data is effective coordination with key stakeholders, particularly when putting in place integrated national systems for collecting administrative data. This approach involves engaging relevant government ministries and a diverse range of entities, including police, health systems, courts, hotlines, shelters and other service providers. One of the first recommended steps here is to create a National or Subnational VAW Administrative Data Coordination Mechanism with leadership, membership, and roles agreed in the Terms of Reference. For examples and further information on how to do this, see Step 2 of UN Women’s guide to Improving the Collection and Use of Administrative Data on Violence against Women.

Where you are supporting primary data collection, it is critical to carefully select interviewers and field supervisors for their ability to handle sensitive issues like violence. They must be thoroughly trained in ethical considerations, empathetic interviewing techniques and the specific challenges of collecting data on violence. This includes understanding some of the contextual nuances affecting responses, handling the emotional distress of respondents and themselves, maintaining confidentiality and safety for all involved, and providing appropriate referral for any interviewees who show signs of distress or request support. Field supervisors also need comprehensive training to support and guide the data collection team effectively. This includes providing ongoing training and following quality control processes to reduce errors relating to possible incorrect sampling, non-response, data processing and coding.

When conducting primary research, it is important to test data collection tools, particularly for large-scale prevalence surveys. Piloting will help test the survey wording, interview training, safety and emotional distress protocols, data capture, quality control, and coding. This stage is also useful for identifying potential issues with response rates and refusals, particularly in subgroups of interest, and ensures the survey’s sample size is adequate for reliable data. The results of the pilot stage are instrumental in finalising the data collection methods and, if weaknesses are identified, doing revisions or even a second pilot before rolling out a survey or scaling up a data collection tool is recommended.

Given the highly sensitive nature of data about violence against women and girls, it is important that robust ethical and safety protocols and procedures are in place to ensure the safety of respondents and the research team. In many cases, this includes securing ethical permissions from national ethics committees and/or university ethics boards. For more information, read how to conduct ethical data collection.