Identifying specific paths that differ PreviousNext
Mplus Discussion > Structural Equation Modeling >
 Jinni Su posted on Friday, September 27, 2013 - 1:20 pm

I am conducting a multi-group analysis (gender as the grouping variable) of to examine a model that includes multiple direct and indirect paths specified. My hypothesis is that all the paths in the model will vary across gender. I compared two models, one with all paths freely estimated across groups and the other with all paths constraint to be equal across groups. chi-square difference test is significant, indicating gender moderation somewhere. My question is, how do I identify which path is need different across gender? I know in AMOS there is a section in the output on critical ratios that help identify the specific difference, where in the Mplus output can I find similar information?
 Linda K. Muthen posted on Saturday, September 28, 2013 - 11:07 am
I would look at modification indices.
 Jinni Su posted on Sunday, September 29, 2013 - 7:49 am
Thank you for your response! I thought about looking at the modification indices too, but then I thought modification indices can only be used as "suggestive" information and we cannot rely on them to conclude anything related to significant difference, right? Maybe I will just need to run a chi-square difference test for each path?

Best regards,
 Linda K. Muthen posted on Monday, September 30, 2013 - 10:46 am
Yes, modification indices can suggest which coefficients you might want to test using chi-square difference testing or MODEL TEST.
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