This hands-on seminar teaches the “nuts and bolts” of advanced regression models. Besides explaining the statistical foundations, the focus of the seminar is on practical applications that include many examples from the area of political sociology and policy studies. The starting point is the classic linear regression OLS estimator and the reasons why this method does not always get the job done. In doing so, we will discuss the topic of causal inference in detail. Subsequently, we will focus on models such as Logistic Regression, Random Effects or Multilevel models, Fixed Effects, Difference-in-Differences, and Structural Equation Models.
Participants will learn how to deal practically with these methods using real data examples in R and Stata. Participants will also have the possibility to discuss their own projects (e.g., Master thesis).
Knowledge of basic linear regression is useful, but not a prerequisite. The course will be held in English.
MA students register via LSF. Doctoral students should register directly with me via conrad.ziller@uni-due.de
TuV: This seminar can be used as a course in the Policy Module. In that case, you must not complete the module with Prof. Behrend's course in the upcoming Winter term (as it builds on her course in the summer), but will have to elect the Policy Analysis course offered by either Prof. Blätte or Prof. Goerres.
Requirements:
- Active class participation
- Readings for each session, which will be digitally available
- Group exercises
- Own analysis and its documentation as term paper
|
