Attrition in panel surveys

April 1st 2015
Universität Mannheim, SFB Seminarraum 016/017 (Erdgeschoss) in L13,17

Prof. Peter Lugtig 

Abstract

Panel studies are popular, mainly because they allo the study of change at both the micro- and macro level. In recent years, many different panel studies have either started or expanded, and this is especially the case in Germany. Sociologists and economists use data from these panel studies to document change within our societies. For example, they study household formation, income dynamics over time, and the link between attitudes and (later) behaviors. One of the biggest problems in panel studies is attrition (dropout). Many respondents stop participating in panel surveys, and they do so for many different reasons. In this talk, I will discuss research documenting the predictors of various types of attrition, and discuss biases that result from this. Does the type of attrition depend on how the panel is managed? In the second part I will discuss potential solutions for reducing dropout in panel studies. Responsive and adaptive designs in particular can be used to tailor the panel management to the specific motivations respondents have for participating and dropping out of the study.

About the speaker

Prof. Peter Lugtig is Assistant Professor at the Department of Methods and Statistics at Utrecht University and Senior Research Officer at the Institute of Social and Economic Research at the University of Essex. He is also editor of "Methodology, Journal of the European Association of Methodology". Among other things, Peter Lugtig regularly contributes to the GESIS Summer School workshops program.