Open huww1998 opened 1 year ago
Hi, if the mediators are independent given the exposure and confounders, the path-specific effect of each mediator can be obtained by fitting a single-mediator model for each mediator separately.
But which modeling approach I can choose ? If I use the weighting-based approach when dealing with parallel multiple mediators by fitting a single-mediator model for each mediator separately, will there be some problems( e.g. outcome model misspecification)?
I am making a multiple mediation analysis( include 4 independent medaitors) and want to obatain path-specific effect of each of medaitors. But there is no available method in the
CMAverse
.