Direct effects in CFA model PreviousNext
Mplus Discussion > Confirmatory Factor Analysis >
 Shang-Min Liu posted on Friday, December 14, 2007 - 11:10 am
I am running CFA model after getting Factor structure for EFA. I am doing step by step follows the handout. My concern is about direct effects between covariates and indicators in MIMIC. How could I test which direct effects would work in my CFA model with covariates?

 Linda K. Muthen posted on Friday, December 14, 2007 - 1:40 pm
If you add a statement like y1-y10 ON x1-x5@0; and ask for modification indices using the MODINDICES option of the OUTPUT command, you can see which direct effect are the most important.
 Shang-Min Liu posted on Friday, December 14, 2007 - 2:14 pm
Thanks for your reply. But after get the result, how to decide how many direct effects need to be include? Do we have any criteria to be reference? Also, would you please let me know the reason to do this way?

Appreciated for your help!
 Bengt O. Muthen posted on Friday, December 14, 2007 - 5:44 pm
You want to include direct effects that are statistically significant, substantively significant, and are interpretable. If you don't include such direct effects your model is misspecified and its estimates distorted.
 Shang-Min Liu posted on Monday, December 17, 2007 - 11:03 am
Thank you for the reply!
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