GMM: Fixing Common Intercepts with Ra... PreviousNext
Mplus Discussion > Growth Modeling of Longitudinal Data >
Message/Author
 Katherine Houghton posted on Wednesday, June 13, 2012 - 10:10 am
I am attempting to build a GMM in which there are 3 known classes and 2 latent classes per treatment group (decided based upon comparison of fit statistics and class sizes). Within each Class, I would like to fix intercepts to be common, but allow the slopes to be random. Intercepts can differ between each class.

I currently have:
%CT#1.C#1%
[i1];
[s1];

%CT#1.C#2%
[i1];
[s1];

etc for CT#2 and CT#3


What do I need to add to my 'I1's to ensure that individuals within a class have common intercepts?

Many thanks,
Katherine
 Linda K. Muthen posted on Wednesday, June 13, 2012 - 1:25 pm
Do you want the means of the intercept and slope growth factors held equal across the known classes or the unknown classes?
 Katherine Houghton posted on Thursday, June 14, 2012 - 1:30 am
Thank you for your reply.

I would like the mean intercepts to be held equal for each latent class within its respective known class. Slopes can be random. That is, I would like the model to find individuals within, e.g. CT1.C1, that have a common intercept, but random slopes. This intercept should not be constrained to be equal to the intercept in, e.g. CT1.C2, CT2.C1, etc.

Would this produce those results:
%CT#1.C#1%
[i1] (1);
[s1];

%CT#1.C#2%
[i1] (2);
[s1];


Or this:

%CT#1.C#1%
[i1] ;


%CT#1.C#2%
[i1];
 Linda K. Muthen posted on Thursday, June 14, 2012 - 11:27 am
I think what you want is:

%ct#1.c#1%
[i1] (1);
%ct#1.c#2%
[i1] (2);
%ct#2.c#1%
[i1] (1);
%ct#2.c#2%
[i1] (2);
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: