Election Studies: A Research Data Management Challenge

Abstract

Election studies present a research data management challenge due to the size and diversity of the data they collect in a concentrated period. We consider how data infrastructure can support election studies to ensure the data they produce is good quality reusable comparative data. Election studies are critical large-scale data investments. When fully realised, election studies provide an unparalleled spatial and temporal resource for studying social attitudes and political behaviour. They have contemporary value for researchers but also as long-term data investments with prospective comparison across time and in comparative cross-national analyses. However, election studies operate under circumstances exceptional to other long-term longitudinal or repeated cross-sectional studies. Elections may occur at irregular intervals or as sudden 'snap' elections. Consequently, studies may have a concentrated period to design and collect data. Here the risk is that data management concerns are relegated to irrelevant or secondary consideration under the pressure to produce publications, with the result that data is not comparable across either time or spatial units. We seek to ensure election studies produce well-documented data, collected in compliance to recognised data harmonisation standards by ensuring data management planning and metadata are incorporated and implemented in the design and execution of the research.