Serial Mediation Analysis Syntax

[Mojtaba Salem]

Dear Stata Users,

My idea was to initial compute a sequential (or serial) moderated mediation in Stata. Because I could not figure out the coding, I decided to divide my sample into high moderator sample and low moderator sample. Now, I need to run a serial mediation analysis with two mediators for each one of the samples:

High Moderator Sample: IV -> Mediator 1 -> Mediator 2 -> DV
Low Moderator Sample: IV -> Mediator 1 -> Mediator 2 -> DV

Here (http://www.ats.ucla.edu/stat/stata/faq/mulmediation.htm) I can find the mediation analysis for two mediators but I think it's not serial mediation. Can anyone give me a hint about the code for serial mediation analysis?

Best,

Mojtaba

[Doug Hemken]

That web page was last updated in 2011. At this point, you might be just as well off to use Stata's "sem" command for this task. You might go to Hayes' website and look for the Hayes and Preacher 2013 citation.

[Phil Bromiley]

I'd like to echo Doug's suggestion. Many mediation tests do not recognize the possibility that errors in the first equation might be correlated with errors in the second equation. sem makes such modelling easier. You might find J. Myles Shaver. 2005. “Testing for mediating variables in management research: Concerns, implications, and alternative strategies,” Journal of Management, 31(3), 330-353. helpful.

[wbuchanan]

You may also want to consider reviewing Hayes, A. (2013). Introduction to mediation, moderation, and conditional process modeling: a regression based approach. New York City, NY: Guilford Press. It provides a fairly thorough coverage of these issues, but the models described by Hayes and Preacher do not have standardized Stata syntax.