32.011 - New Weighting Methods for Causal Mediation Analysis

Session Type: Professional Development Course

Abstract

Many important research questions in education, prevention science, and social sciences relate to how interventions work. Alternative theories often provide competing explanations for the causal mechanisms, that is, the processes through which an intervention succeeds or fails. A theoretical construct characterizing the hypothesized intermediate process is called a mediator. Conventional methods for mediation analysis generate biased results when the mediator-outcome relationship depends on the treatment condition. These methods also tend to have a limited capacity for removing confounding associated with a large number of covariates. This workshop teaches the ratio-of-mediator-probability weighting (RMPW) method for decomposing total effects into direct and indirect effects in the presence of treatment-by-mediator interactions. RMPW is easy to implement and requires relatively few assumptions about the distribution of the outcome, the distribution of the mediator, and the functional form of the outcome model. We will introduce the concepts of causal mediation, explain the intuitive rationale of the RMPW strategy, and delineate the parametric and nonparametric analytic procedures. Participants will gain hands-on experiences with a stand-alone RMPW software program that eases computation and facilitates users’ analytic decision-making. We will also provide SPSS, SAS, Stata, and R code for interested users. The target audience includes graduate students, early career scholars, and advanced researchers who are familiar with multiple regression and have had prior exposure to binary and multinomial logistic regression. Prior knowledge of causal inference is not required but will be a major plus. Each participant will need to bring a laptop for hands-on exercises.

Sub Unit

• Professional Development and Training Committee

Instructors

• Guanglei Hong, University of Chicago

• Jonah Deutsch, University of Chicago