Acceso usuarios
Martin Kroh, Florin Winter, Jürgen Schupp
Panel conditioning has posed one of the main challenges to panel studies since their inception in the social sciences. Aside from the risk of reactivity to previous interviews, there is reason to expect that cumulative survey experience increases the reliability of data emanating from panel studies relative to cross-sectional surveys. This positive aspect of recurrent interviewing for data quality has been given relatively little attention in the empirical research to date. Drawing on observational data from 30 waves of the German Socio-Economic Panel (SOEP), we study the effect of individual survey experience on reliability, focusing on person-fit statistics from item-response models. The analysis documents that four years of survey experience produce a higher increase in person reliability than tertiary education compared to primary education.
