Survey sampling in wartime: addressing challenges for cross-national and longitudinal studies

Conflict situations present unique challenges for survey sample design, three of which are examined in this paper within the context of the ongoing full-scale war in Ukraine. Firstly, defining the population becomes a complex task as a result of substantial population changes. This includes internal displacement, a significant proportion of the population emigrating with potential future return, and a considerable number of individuals actively engaged in combat who are currently unreachable for interviews. Secondly, defining households, particularly for longitudinal studies, is complicated by the temporary separation of many families. Thirdly, accurate accounting for vacant and demolished dwellings is essential, as national statistics often lack precise and reliable data in these areas. This study outlines the design of two samples: one for a cross-national European Social Survey in Ukraine (ESS) and another for the longitudinal household panel study, UKRAINS. By carefully addressing these complex sampling challenges, it is possible to develop high-quality probability samples that account for population mobility and unpredictability in a wartime context.