The central role of the mediator process in mediation analysis
Keywords: Mediation, Causal inference, Time-varying confounding
TL;DR: In mediation analysis, failure to account for the underlying mediator process has profound implications for defining the relevant causal estimand of interest as well as its identification and estimation.
Abstract: Causal mediation has traditionally been framed as the effect of an exposure on an outcome through some intermediate variable, where each variable is measured at three sequential time points. However, definitions of mediated effects and their corresponding identification assumptions generally ignore the fact that the mediator of interest is, in many if not most circumstances, a stochastic process indexed by time from baseline to follow-up. I demonstrate that the failure to account for the mediator process has profound implications for defining the relevant causal estimand of interest as well as its identification and estimation. Additionally, I introduce novel versions of direct and indirect effect definitions that account for the entire mediator process.
Submission Number: 32
Loading