Shared Cause Adjusted Mediation (SCAM): Introduction of method with empirical applications

A genuine mediated effect, also called an indirect effect, is the part of the relationship between a predictor (X) and an outcome (Y) that operates through a third variable, called a mediator (M). However, estimated mediated effects do not prove genuine mediation as they only require variables to be correlated and correlations may be due to confounding. Here, I propose adjusting for effects that are expected if the predictor, outcome, and mediator(s) share a common latent cause. In an application of this method on four studies using serial mediation analyses, only 5 of 24 (21%) reported effects remained statistically significant after such adjustment for a shared latent cause.