This course is designed to introduce students to the social science and statistical literature on causal inference. Being a half-semester course, it necessarily provides selective coverage of key ideas in the voluminous literature on causation and causal inference. This class is differentiated from the Causal Inference I course in Political Science by its focus on the work of Judea Pearl and related approaches. Topics to be covered include: directed acyclic graphs and causal identification, the back-door criterion, the front-door criterion, instrumental variables, and relationships to the Neyman-Rubin framework. Prerequisites are POLSCI 599 and POLSCI 699, the equivalent, or permission of the instructor. In addition, a good grasp of the R language for statistical computing will be extremely useful for the problem sets.
Intended Audience:
Graduate students in social science disciplines
Class Format:
Lecture