Re: st: Sign. Test for indirect effects x-->m1-->m2-->y

["Casey P. Durand" <durandca@usc.edu>]

Alan,

I believe the Sobel test can be extended to the situation you describe
with multiple mediators (see reference below for more detail), though
as you allude to, the Sobel test is a less than satisfactory test in a
mediation context.  Bootstrapping would be ideal, though I can't offer
much advice on running the multiple mediator analysis in Stata.  I
personally use Mplus for any situation like that, especially for the
bootstrapping capabilities.


Tests of the Three-Path Mediated Effect. Aaron B. Taylor, David P.
MacKinnon and Jenn-Yun Tein. Organizational Research Methods 2008 11:
241

Casey

On Thu, Mar 3, 2011 at 11:47 AM, Alan Acock <acock@mac.com> wrote:
>
> I hope this is a simple question. The Sobel test is okay for x-->m-->y models with a single mediator. The sureg and nlcom command is great for two parallel mediators x->m1-->y and x-->m2-->y,
> I can't figure out how to test it with m2 mediates m1 which mediates x as in
> x-->m1-->m2--y.
>
> --Alan Acock
> *
> *   For searches and help try:
> *   http://www.stata.com/help.cgi?search
> *   http://www.stata.com/support/statalist/faq
> *   http://www.ats.ucla.edu/stat/stata/

*
*   For searches and help try:
*   http://www.stata.com/help.cgi?search
*   http://www.stata.com/support/statalist/faq
*   http://www.ats.ucla.edu/stat/stata/