topic Can CAUSALMED procedure handle more than one mediator variables? in Statistical Procedures

Can CAUSALMED procedure handle more than one mediator variables?

joe66 — Thu, 07 Jun 2018 21:23:20 GMT

Hi , I got one mediator variable that has three levels, such as "difficult" , "medium" and "easy", I know CAUSALMED can handle binary or continuous mediators, however I am not sure if I can include two binary mediators in the model by creating two dummy variables for "medium" and "easy"?

Is there any way I can apply this three level categorical mediator?

Thanks,

Robin

Re: Can CAUSALMED procedure handle more than one mediator variables?

StatDave — Fri, 08 Jun 2018 13:57:25 GMT

The methodology that CAUSALMED employs does not allow for more than one mediator. However, see the second example in this note.

Re: Can CAUSALMED procedure handle more than one mediator variables?

joe66 — Fri, 08 Jun 2018 14:16:43 GMT

Thanks for reply. The example involve two mediators, but I noticed that it does not include bootstrapping as Causalmed, how could I include bootstrapping in this method?

Robin

Re: Can CAUSALMED procedure handle more than one mediator variables?

tamhane — Thu, 18 Oct 2018 15:47:40 GMT

Refer to this paper: "Mediation Analysis Allowing for Exposure–Mediator Interactions and Causal Interpretation: Theoretical Assumptions and Implementation With SAS and SPSS Macros" by Valeri and VanderWeele.