Direkt zum Inhalt

Attrition and Selectivity of the NEPS Starting Cohorts: An Overview of the Past 8 Years

Weitere referierte Aufsätze

Sabine Zinn, Ariane Würbach, Hans Walter Steinhauer, Angelina Hammon

In: AStA Wirtschafts- und Sozialstatistisches Archiv (2020), 14, S. 163–206


This article documents the number of target persons participating in the panel surveys of the National Educational Panel Study (NEPS) as well as the number of respondents who temporarily dropout and of those leaving the panel (attrition). NEPS comprises panel surveys with six mutually exclusive starting cohorts covering the complete life span. Sample sizes, numbers of participants and temporary as well as final dropouts and participation rates are reported in detail for each wave and for subsamples, if applicable. Sample particularities, such as the conversion of temporary dropouts into final ones, are elaborated on. All figures presented are derived from the corresponding Scientific Use Files (SUFs) published by February 1, 2018. Selectivity due to attrition (i.e., final dropouts) is studied. For this purpose, we examine how attrition distorts the NEPS samples with respect to relevant design variables (such as stratification criteria) and panel member characteristics (like sex and birth year). In detail, we study the panel status of each panel member, that is being part of the panel or having dropped out finally, along all of the panel waves with respect to starting cohort and population specific characteristics. We conclude this article with some recommendations for dealing with the detected selection bias in statistical analyses.

Angelina Hammon

Research Associate in the German Socio-Economic Panel study Department

Sabine Zinn

Vice Director SOEP and Division Head Survey Methodology and Management in the German Socio-Economic Panel study Department

Hans Walter Steinhauer

Survey Statistician in the German Socio-Economic Panel study Department

Keywords: National Educational Panel Study (NEPS), Case numbers, Selectivity, Attrition, Discrete time event history model

Frei zugängliche Version: (econstor)