Class specific constraints in LCGA PreviousNext
Mplus Discussion > Growth Modeling of Longitudinal Data >
Message/Author
 George Blanze posted on Thursday, August 13, 2009 - 7:27 am
I'd like to constrain my first class to follow a linear trajectory, while allowing class 2 & 3 to have a cubic trajectory. What is the syntax for this? Right now I have:

Analysis:
type = mixture ;
Model:
%overall%
i s q cu | t1@0 t2@1 t3@2 t4@3 t5@4;
c ON x1 x2 x3;

Thank you for your help.
 Linda K. Muthen posted on Thursday, August 13, 2009 - 10:59 am
Model:
%overall%
i s q cu | t1@0 t2@,1 t3@,2 t4@,3 t5@,4;
c ON x1 x2 x3;
%c#1%
[q@0 cu@0];
q@0 cu@0

I would also divide the time scores by 10 as shown above because the cubic scores may get large.
 George Blanze posted on Thursday, August 13, 2009 - 11:39 am
In order to use the extra lines you listed, I need to add 'algorithm=integration' to the analysis step. When I do this, I receive a "memory" error. Is there a way to do this without numerical integration? Or do you have any recommendations for solving the memory error?

Thank you.
 Linda K. Muthen posted on Thursday, August 13, 2009 - 2:14 pm
Numerical integration is required for this analysis if the outcome is categorical. You can try adding INTEGRATION=7; to the ANALYSIS command.
 George Blanze posted on Monday, August 17, 2009 - 8:29 am
"I would also divide the time scores by 10 as shown above because the cubic scores may get large."

Would you mind explaining the reason for this? I was under the impression that the scales of the variables in a multiple regression were not important.

Thank you.
 Linda K. Muthen posted on Monday, August 17, 2009 - 10:08 am
Scales affect the efficiency of the optimization algorithm.
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: