Indirect effects PreviousNext
Mplus Discussion > Structural Equation Modeling >
 Jeremy Miles posted on Thursday, May 21, 2009 - 4:33 pm

I'm running some (relatively) straightforward mediation models, with a categorical outcome.

I run a model with a predictor of the outcome:

outcome on predictor1 predictor2;

And then I add a couple of mediators to the model, and run it again, with the model indirect: option -

outcome on predictor1 predictor2 mediator1 mediator2;

mediator1 mediator2 on predictor1 predictor2;

mediator1 with mediator2;

outcome ind predictor1;
outcome ind predictor2;

I would have thought that the total effect from the indirect command was equal to the effect for the outcome on the mediator for the first model, but it's not. It's similar - 0.095 vs 0.105.

Is there an obvious reason that I'm missing why not?

(there are a couple of control variables in the model as well. the model has 0 df).


 Linda K. Muthen posted on Thursday, May 21, 2009 - 5:27 pm
Please send the output and your license number to
 Jeremy Miles posted on Friday, May 22, 2009 - 11:28 am
Oops. I'd left estimator=ML in one of the files. That explains it.

Thanks for making me look again.
Back to top
Add Your Message Here
Username: Posting Information:
This is a private posting area. Only registered users and moderators may post messages here.
Options: Enable HTML code in message
Automatically activate URLs in message