A/HRC/7/6 page 13 • Surveys constructed to enable analysis across multiple definitions, including recognized international standards • Both recent (12 months) and lifetime time frames • Representative and inclusive sampling frames: both sufficient size to make findings reliable and efforts to reach/include vulnerable groups • Measurements of incidents, frequency and severity, where relevant • Strong ethical practice 46. Recognized ethical standards for prevalence surveys can be summarized as follows:18 • Interviews conducted in a safe and confidential setting • Confidentiality protocols, including anonymizing of all stored and published data • Well-trained female interviewers, with debriefing opportunities • Careful construction of questions and instruments to take interviewees into and out of explicit discussions of violence • Access to post-survey support/advice for those who may need it 47. Prevalence surveys should also routinely collect data on: • Basic demographics • Relationship to perpetrators • Life contexts in which violence against women takes place • Reporting and help-seeking, and barriers to it • Forms of harm and impact 48. This final category needs to extend from immediate and health-based measures to cover how violence narrows women’s and girls’ enjoyment of fundamental rights and freedoms in public and private life, including with respect to social inclusion and livelihoods. Both lifetime and 12-month measures are commended: the former ensures that serious violence occurring in childhood and early adulthood, with lifetime consequences, is included; the latter provides a 18 Co-ordination Action on Human Rights Violations, “Perspectives and standards for good practice in data collection on violence against women”. Paper for Data collection as a prerequisite for effective policies to combat violence against women conference, CoE, Lisbon, 5 July 2007.