No title
This article provides an overview of the role mediation analyses play in causal inferencemethodology and in the estimation of the effect an exposure may have on an outcome throughdirect and indirect pathways. We discuss traditional approaches to mediation and where thesemay lead to bias, before discussing more advanced methods that can be used to overcome thesebiases in situations with single and
