Latent distal/ggmm regression with kn... PreviousNext
Mplus Discussion > Growth Modeling of Longitudinal Data >
Message/Author
 Keri Jowers posted on Saturday, March 21, 2009 - 1:35 pm
good afternoon,

i'm trying to determine whether the association between a latent distal and a ggmm differ by gender using knownclass. the ggmm differs by gender, so there are 2 separate models for that piece. when i run the ggmm and knownclass piece (the 1st 4 combinations), it works and teh results make sense. how do i correctly add in the latent distal? any wisdom you can offer would be much appreciated! (in my actual input, i have set the starts for each according to the original gmm/lca model)

Variable:
names are ...
classes = cg (2) cgmm(2) clca(4);
knownclass = c (g = 0 g = 1);

Analysis: type is mixture;

Model:
%overall%
i s | y1@0 y2@1 y3@2 y@3;
cgmm#1 on cgender#1;
clca#1 on cgender#1;
clca#2 on cgender#1;
clca#3 on cgender#1;
%cgender#1.cgmm#1%
%cgender#1.cgmm#2%
%cgender#2.cgmm#1%
%cgender#2.cgmm#2%
%cgender#1.clca#1%
%cgender#1.clca#2%
%cgender#1.clca#3%
%cgender#1.clca#4%
%cgender#2.clca#1%
%cgender#2.clca#2%
%cgender#2.clca#3%
%cgender#2.clca#4%
 Bengt O. Muthen posted on Saturday, March 21, 2009 - 2:47 pm
The distal can simply be included in the usev list without saying anything in the model - the default is that the distal means/thresholds vary across all classes.

For more specific response, send your input, output, data, and license number to support@statmodel.com.
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: