Skip to content!

July 12, 2018


Introduction to the new SOEPlong-version. From raw data to longitudinal analyses in 3 hours
Post-conference workshop 2018 SLLS Conference


July 12, 2018

The German Socio-Economic Panel Study (SOEP) is a wide-ranging representative longitudinal study of private households in Germany, located at the German Institute for Economic Research, DIW Berlin. Some of the many topics include household composition, occupational biographies, marital biographies, employment, earnings, health and subjective indicators. The panel was started in 1984 and offers longitudinal information for more than 80.000 persons in about 20.000 households.

The purpose of the workshop is to introduce and familiarize doctoral students as well as senior researchers with the file structure and potential of the SOEP data. Specifically, we explore the potential of “SOEPlong”, a new mode of data-provision based on datasets in long-format. Particularly, the workshop focuses on the longitudinal components of the data. We demonstrate and discuss the ability of the longitudinal household design to a) validate causal interpretations, b) combine characteristics from different phases of the life course and c) track dynamics between partners (and other household members) over time.

Given the huge amount of data (and the correspondingly large number of files) covering 30 years of micro data at the household, individual and event levels, the SOEP dataset is growing into an increasingly complex structure which can be difficult to handle for users, even if they are experienced empirical analysts. The workshop shall serve to help users cope with this complexity. Additionally, the potential of longitudinal data in general (and the SOEP in particular) to model life-course dynamics in sociology is still largely underused. Therefore, the course aims at encouraging the sociological research community to make stronger use of the benefits of this type of data in research projects on dynamics of objective, subjective and biographical indicators.