R-Square for direct and indirect effects PreviousNext
Mplus Discussion > Structural Equation Modeling >
 skye woestehoff posted on Monday, April 21, 2014 - 4:27 pm

I ran a latent variable path analysis with three IVs, 2 mediators, and 1 DV. The "standardized" command gets you the R-square for each DV. What I am trying to find is the R-square/percent of variance explained for each effect, both direct and indirect.

For example, is there a way to find the R-square for IV 1 on the DV (both directly and indirectly through one of the mediators)?

Thank you!
 Linda K. Muthen posted on Tuesday, April 22, 2014 - 10:10 am
R-square cannot be split up because the covariates are correlated.
 skye woestehoff posted on Wednesday, May 07, 2014 - 11:39 am
Hello again,

I'm still having a bit of trouble. Is there a way to get the percent of variance explained for the direct and indirect effects? I know you can get to that using the standardized residual variances. However, I am having trouble finding out how to get the residual variances for the paths in my model (I've only been able to find the residual variances for my two mediators and not the direct and indirect effects of my IVs on my DV). Is that possible to do?
 Linda K. Muthen posted on Wednesday, May 07, 2014 - 1:48 pm
Each dependent variable has its own R-square. There is no R-square for the indirect effect.
 swoestehoff posted on Wednesday, May 07, 2014 - 1:52 pm
What about residual variances? Do those exist for indirect effects?
 Linda K. Muthen posted on Wednesday, May 07, 2014 - 3:14 pm
Each dependent variable has its own residual variance. There is not one for an indirect effect just for the dependent variables that are part of it.
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