Uses of Graphical Causal Models in the Social Sciences - Einzelansicht

The seminar discusses practical uses of graphical causal models for the social sciences. Particularly it will discuss at length the use of "Directed Acyclic Graphs" -- which can be seen as a nonparametric generalization of classical path diagramms. The seminar starts by replicating the concept of causality, path diagrams, non-parametric models. It then goes on discussing the features of DAGs in detail. In the last part of the seminar we will read published papers of the social sciences using the tool of DAGs to understand (and criticize) their contents.

Elwert, Felix (2013). Graphical Causal Models. Pg. 245--273 in Morgan, S.L. (Ed.), Handbook of Causal Analysis for Social Research. Springer: Dortrecht.

Elwert, Felix and Winship, Christopher (2014). Endogenous Selection Bias: The Problem of Conditioning on a Collider Variable. Annual Review of Sociolgy 40, 31--53.

Participants should have passed the modul "Methoden der empirischen Sozialforschung" (i.e. Applied Regression Analysis and Research Design)

For passing: Homeworks

Modul examination: Seminar paper