Is My Panel Survey Still Representative? A New Method for Detecting whether Two Subsamples Represent the Same Universe--The German Socio-Economic Panel Study Experience

• Activity Number: 279
• Type: Contributed
• Date/Time: Tuesday, August 10, 2004 : 2:00 PM to 3:50 PM
• Sponsor: Section on Survey Research Methods
• Author(s): Jan Goebel*+ and Joachim R. Frick and Edna Schechtman and Gert G. Wagner and Shlomo Yitzhaki
• Companies: DIW Berlin and DIW Berlin and Ben Gurion University and DIW Berlin and Hebrew University and Central Bureau of Statistics
• Keywords: panel studies ; Gini decomposition

Abstract

A most important shortcoming of panel surveys is bias arising from selective attrition. Based on data of the German Socio-Economic Panel Study (SOEP) we analyze potential artifacts (level, structure, inequality of income) by comparing results for two independently drawn panel subsamples started in 1984 and 2000, respectively. Both samples are carried on using the same set of follow-up rules. We apply ANOGI (Analysis Of Gini) techniques, the equivalent of ANOVA (Analysis Of Variance) performed with Gini coefficient. The decomposition followed is presented in Yitzhaki (1994). We rearrange, reinterpret, and use the decomposition in the comparison of subpopulations from which the different samples were drawn. Taking into account indicators for income--and for control purposes those for education and satisfaction as well--significant differences between these two samples with respect to income inequality are found in the first year, which start to fade away in wave two and disappear in wave 3. We find credible indication for these differences to be driven by changes in response behavior of short-term panel members rather than by attrition among members of the longer-running sample.