How to fix thresholds to set latent m... PreviousNext
Mplus Discussion > Categorical Data Modeling >
Message/Author
 laura francesca scalas posted on Wednesday, May 28, 2014 - 3:46 am
Greetings,
In a single group CFA model using WLSMV estimation with ordered categorical items (not binary, we have over 3 categories), how would you recommend fixing the thresholds in order to set the metric for the latent means (i.e. to freely estimate means)? (e.g., with continuous indicators, the intercept of a reference item can be fixed to 0, what about categorical indicators?)
 Linda K. Muthen posted on Wednesday, May 28, 2014 - 2:16 pm
You can fix the first threshold and estimate the mean.
 laura francesca scalas posted on Thursday, May 29, 2014 - 6:59 am
Hi,
just a clarification. Do I need to fix the first threshold to 0 or to 1?
Thanks
 Bengt O. Muthen posted on Thursday, May 29, 2014 - 1:13 pm
Doesn't matter in principle, but fixing it to 0 is more common.
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: