Local independence assumption and cou... PreviousNext
Mplus Discussion > Latent Variable Mixture Modeling >
Message/Author
 John G. Orme posted on Tuesday, December 13, 2011 - 12:49 pm
Is it possible with Mplus to relax the local independence assumption when using count variables to estimate a latent class model?

Thanks for any help you can provide with this!
 Linda K. Muthen posted on Wednesday, December 14, 2011 - 11:54 am
You can add covariances among the count variables. You cannot use the WITH option to do this. You can use the BY option as follows. Note that each covariance is one dimension of integration so adding many will result in a computationally demanding analysis.

f BY u1@1 u2;
f@1; [f@0];

The covariance parameter is found in the factor loading for u2.
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: