Multiple group analysis for semiconti... PreviousNext
Mplus Discussion > Categorical Data Modeling >
 Scott Weaver posted on Friday, May 19, 2006 - 10:19 am
I wish to test a semicontinuous model (single time point; not a growth model like the example in the Mplus Manual) and compare the regression parameters across multiple groups. With semicontinuous models, ML with numerical integration is required (when a latent variable is predicting a categorical variable); thus I can't use the GROUPING IS statement. So I tried using the KNOWNCLASSES statement, but I am given the error message that I need to specify TYPE=MIXTURE. When I specify TYPE=MIXTURE, the output indicates that I need a categorical latent variable - which I don't have.
Can Mplus 4.1 handle multiple group modeling for semincontinuous models? If so, I can I coax the program into running?

Thank you!
 Linda K. Muthen posted on Friday, May 19, 2006 - 1:38 pm
Just say CLASSES = c (1); in the VARIABLE command.
 Linda K. Muthen posted on Friday, May 19, 2006 - 8:14 pm
Actually, when you use the KNOWNCLASS option, you need to use the CLASSES option as shown in Example 7.21. You will have one variable instead of 2 on the CLASSES statement.
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