Oliver Lipps has been working on methodological issues in social science panel surveys since 1995 and is currently a member of the Swiss Household Panel team (SHP). He has published on data quality in cross-sectional and longitudinal designs, mostly using multilevel modeling tools.
Ursina Kuhn is a member of the Swiss Household Panel team. Her research interests are applications using household panel from Switzerland, Germany and Great Britain, especially in the domains of political behaviour and income.
This course introduces the structure, data management, and analysis of panel data in general and the SHP in particular. Participants will learn to understand the structure of (household) panel data, to prepare the data for panel analyses, and to apply descriptive statistics and analytical models. After a detailed presentation of the SHP, we will learn how to organise data for different types of research questions which are specific to that type of data. This includes for example creating longitudinal data files, partner or parent-children data files, and spells and transitions. We will address issues of nonresponse and weighting. Analysis techniques include cross-tabulations, regressions (fixed effects, random effect, multilevel) and, if time allows, event history analysis. One part of the practical work will be that participants will develop their own small research project using SHP data. There will be time to solve individual problems arising.
(Kind of material covered during class)
Experience in statistical analysis of cross-sectional (survey) data is required. This includes experience with statistical software (such as SPSS, Stata, SAS) and a basic knowledge of cross tabulations and regressions. In the course, Stata will be used. Experience with Stata is useful but not required.