SOCIOL2245: Causal Inference for the Social Sciences

Semester: 

Spring

Offered: 

2021
It is an adage that association does not imply causation, yet causality is at the center of most social science inquiries. This course introduces a set of concepts, assumptions, and methods that allow students to rigorously assess causal relationships from experimental and observational data. Drawing on applications from sociology, economics, and political science, we discuss a variety of research designs and statistical methods, including randomized experiments, regression adjustment, matching and weighting, instrumental variables, regression discontinuity designs, panel data models (including difference in differences), causal mediation analysis, and nonparametric/machine learning methods. The class will be a mixture of lectures, discussions, and computer work.