300109 Social contexts in the life course: analyzing multilevel-data (S) (WiSe 2015/2016)

Individuals are embedded into multiple social contexts throughout their life course, for instance into families, schools, neighborhoods, and work organizations. The attempt to understand and explain individual behavior and its consequences without taking into account this embeddedness is at best insufficient.

Today, many data sets used in the social sciences contain information about how actors are embedded into social contexts. So called multilevel models (also known as hierarchical linear models and mixed models) have become a common statistical tool for the analysis of clustered or grouped data. Their advantages are obvious: instead of treating observations incorrectly as unrelated, they explicitly take the clustering of observations into account and allow for modeling how characteristics of the higher level impact units at the lower level – for example, how characteristics of neighborhoods affect their residents or how characteristics of schools affect students.

The goal of this seminar is to develop theoretical and practical knowledge on actors’ embeddedness into social contexts and apply this knowledge using methods of multilevel modeling. The first part of the seminar will cover basic theoretical models of action and social embeddedness, drawing on life course research and rational choice theory. In the second part of the seminar we will jointly work out and apply methods of multilevel analysis.

Requirements for participation, required level

Basic knowledge of statistics and Stata are required.

Bibliography

Hedström, P., 2008, Dissecting the Social

Galster, G. C., 2012, The mechanism (s) of neighbourhood effects: Theory, evidence, and policy implications. In Neighbourhood effects research: New perspectives (pp. 23-56). Springer Netherlands.

Rabe-Hesketh, S., & Skrondal, A., 2012, Multilevel and longitudinal modeling using Stata.