Causal Mediation Analysis

Session Submission Type: Course

Description

Causal mediation analysis lies at the very heart of scientific inquiry. It seeks to uncover not just whether but also why an exposure, treatment, or stimulus affects an outcome by quantifying the processes and mechanisms through which a causal effect operates. That is, it aims to identify causal chains that connect an exposure to an outcome via intermediate variables known as mediators. This class will cover methods for analyzing causal mediation with an emphasis on social science applications. It will use precise notation and accessible conceptual diagrams to lead students from basic definitions of effects, via minimally necessary assumptions, to cutting-edge estimation procedures. It will provide a comprehensive guide for analyzing causal mediation using modern techniques, including effect decomposition, adjustment for confounding, analysis of multiple mediators, and estimation via regression modeling, inverse probability weighting, and machine learning methods. The class will address both theory and conceptual material alongside practical implementation. The course will draw on the instructor's new book from Cambridge University Press, Causal Mediation Analysis, and its associated R and Stata packages.

Sub Unit

• Courses

Presenter

• Geoffrey Thomas Wodtke, University of Chicago

Session Organizer

• Geoffrey Thomas Wodtke, University of Chicago