|
Abstract:
|
Mediation analysis explores the degree to which an exposure's effect on an outcome is diverted through a mediating variable. Here we introduce and extend the classical regression framework for conducting mediation analysis in which estimates of causal mediation effects and their variance are obtained from the fit of only a single regression model. The vector of changes in exposure pathway coefficients, the so-called Essential Mediation Components (EMCs), is used to estimate causal mediation effects. Because mediation effects are simple functions of the EMCs, an analytical expression for their model-based variance follows directly. The approach is extended to non-nested mediation systems with a "double-sweep". Furthermore, we propose new visualizations for mediation effects and we explain why estimates of the total effect may depend on the framework in which it is estimated. Using data from social science studies, we also provide extensive illustrations of the usefulness of this framework and its advantages over popular approaches.
|
