The design and conduct of surveys should always be calibrated to specific circumstances and goals. It is, therefore, not possible to give general guidance on what surveys should address and how the questions should be framed. However, a general sense of how transitional justice consultation surveys appear can be garnered from a brief review of two recent ones. However, it should be kept in mind, as discussed further below, that the construction of survey questions is a task for the experts. Kosovo (2007). A survey conducted by the United Nations Development Programme focused on public opinion on human rights standards, violations and accountability; explored ethnic bias in the assessment of war crimes and missing persons; addressed challenges faced by judges and prosecutors; and sought to identify appropriate forms of reparation. The questions addressed, among other topics, how many people of various ethnicities had experienced human rights violations; levels of support for the resolution of problems of missing persons; achievement of reconciliation among ethnic communities; and degrees of preference for such forms of reparation as material compensation, rehabilitation and formal recognition of victim status. Northern Uganda (2007). The regional survey conducted by the Berkeley-Tulane Initiative on Vulnerable Populations and the International Center for Transitional Justice sought to describe attitudes to peace and justice. The questions addressed, among other topics, levels of exposure to various forms of violence (such as abductions and damage to property); prioritization of demand for such basics and services as health care, food and justice; preferences among such forms of reparation as compensation, apologies and reconciliation; and preferences between traditional and formal justice mechanisms. A drawback of the survey method is the manner in which it tends to propose a limited range of options, to which the respondents react. This may preclude respondents from suggesting novel or otherwise unanticipated transitional justice arrangements. Surveys that are not carefully constructed to overcome this limitation may reduce policymakers’ opportunity to take into account forms of traditional or local practice that might play an important role within a transitional justice programme. Qualitative data methods can address some of these risks. B. Qualitative consultations While quantitative consultations are based on the collection of data in the form of numbers, qualitative consultations are concerned with collecting data in the form of words. Those data are typically unstructured and statistical methods cannot be used to analyse them. Whereas, in quantitative consultations, structured interviews and questionnaires reflect the researcher’s interest and concerns, qualitative research techniques focus on the respondents and seek to gain an in-depth understanding of their personal views, attitudes and beliefs on certain issues. The aim of the researcher is to view the issue through the eyes of the respondent. 8