MESSAGE: THE LATENT VARIABLE COVARIANCE PreviousNext
Mplus Discussion > Confirmatory Factor Analysis >
Message/Author
 Stella Summer posted on Tuesday, April 03, 2012 - 8:06 am
Linda,

I've got a warning message saying that:
THE LATENT VARIABLE COVARIANCE MATRIX (PSI) IN GROUP GIRLS IS NOT POSITIVE DEFINITE...

I have detected the problem, two latent factors had a std-solution with correlation above 1. Could I constrain this factor's variance to 1 (FACTOR1@1;) to overcome this problem? Or are there other approches available?


Thank you,
 Linda K. Muthen posted on Tuesday, April 03, 2012 - 10:09 am
If two factors correlate greater than one, the model needs to be changed for girls. The first step in a multiple group analysis is to fit the model in each group separately. If the same model does not fit in each group, comparisons across groups do not make sense. You might want to start with an EFA in each group to see if the same number of factors fits in each group.
Back to top
Add Your Message Here
Post:
Username: Posting Information:
This is a private posting area. Only registered users and moderators may post messages here.
Password:
Options: Enable HTML code in message
Automatically activate URLs in message
Action: