S Multilevel Models (AMS, Teil I / II)

Social scientists are often confronted with hierarchical structured data: Textbook examples include students grouped into classes (belonging to schools belonging to geographical region) as well as individuals living in neighbourhoods (grouped into cities grouped into regions) or, in a comparative research perspective, individuals grouped into countries. Theoretical models in these settings often assume cross-level interactions between the individual level and higher levels. A common assumption is that the social composition of a school has an effect on the individual student performance or that the neighborhood context influences the individual probability of delinquent behavior. Statistical models referred to as multilevel (linear) models, mixed-effects models, covariance component models or random-effects models have been proposed in the literature for this kind of data and are often rated superior to simple OLS regression. The course will cover an introduction into practical application and interpretation of multilevel models for a range of different data structures. In addition to computer exercises, research examples and scientific papers using multilevel analysis in different fields will be discussed. Please note that the course will be held in English.

Voraussetzungen für Studiennachweise / Modulprüfungen: Modulprüfung: active participation, completion of exercises and term paper. Studiennachweis: active participation, completion of exercises.


Sebastian Gerhartz


  • Montag, 01.04.2019 (1. Termin)
    14:00 bis 16:00 Uhr
    GD E2/208 CIP-Pool


Bitte melden Sie sich in eCampus für die Veranstaltung an.