Berkeley Causal Inference
Workshop 2020 - CANCELLED*

Mechanisms for Social Science:  Applied Tools for Causal Mediation Analysis

August 5-7, 2020
University of California, Berkeley

*A Note from Our Team (March 29, 2020)

Like many of you, COVID-19 has shaken up our community in ways we did not anticipate. After deep consideration and guidance from UC Berkeley Facilities Services, we have decided to cancel the Berkeley Causal Inference Workshop scheduled for August 5-7, 2020.  We sincerely apologize for any inconvenience but must prioritize the safety and well-being of our attendees, workshop speakers, planning team, and greater community. 

We look forward to offering high-quality learning experiences to social scientists in the future. Be safe and stay positive! 

About the Workshop

Mediation analysis investigates the direct and indirect pathways (“mechanisms”) that connect causes to effects. Mediation analysis is central to testing social theories and evaluating interventions. This 3-day workshop offers social scientists and applied researchers an applied introduction to causal mediation analysis. Using real research examples, this hands-on workshop develops intuition, builds analytic skills, and empowers participants to apply the tools of modern causal mediation analysis to their own research.

The Speakers


Our award-winning faculty offer 20 years’ combined experience teaching causal inference to social scientists around the world. Felix Elwert (University of Wisconsin), David Harding (UC Berkeley), and Geoff Wodtke (University of Chicago).

 Program Overview

The workshop introduces the tools of modern causal mediation analysis with lectures and many hands-on guided exercises in Stata and R. Participants will learn about new classes of mediation questions (“estimands”), potential outcomes, graphical models, assumptions, linear and categorical models with or without interactions, dynamic models, and unobserved confounding.

Program Topics will include: 






This intensive 3-day workshop will introduce causal mediation analysis, potential outcomes and graphical methods, explain the estimands, estimators, and assumptions of modern mediation analysis, and use hands-on guided exercises in R and Stata.