How do you analyze mediation with several treatments in R

Question

I would like to analyze the data from the experiment using the intermediary analysis in R. However, the experimental design is a full factorial design in three variables (two continuous, one categorical), and I can not find an explanation of how to implement mediation in R with several procedures. I read the documentation in the mediation package, but they don't seem to provide ways to extend X beyond a single call. Similarly, I cannot find a way to do this in MBESS or lavaan .

I found a very recent article discussing the statistical theory / approaches needed to implement multiple treatments in mediation analysis, Hayes and Preacher 2014 ( http://quantpsy.org/pubs/hayes_preacher_2014.pdf ), but unfortunately they are give only codes for the implementation of their approach in Mplus, SPSS and SAS. I need to implement this in the next couple of days for a presentation, so I don’t have time to speed up in another program, to do this, I need to implement it in R.

Does anyone know if there is an implementation in R that I missed? Or if there is a way to implement this approach outside the package?

(I understand that I can transform my full factorial design into a single processing, considering each combination of 3 factors as a level, but such an analysis would not help.)

Answer 1

After a more comprehensive search, it turned out that there is currently no R package that can handle intermediation with multiple processes (i.e. predictors). Packages in R that can mediate include: mediation , MBESS , lavaan , multimed , bmem and OpenMx . However, it appears that those that process multiple predictors (such as OpenMx ) also require a model encoded as a RAM path model. This is the type of model needed for a full path analysis code and, in particular, more complex than the simplified coding of the mediation model (for example, in mediation ) that I was hoping for.