Gmm with distal outcomes PreviousNext
Mplus Discussion > Latent Variable Mixture Modeling >
Message/Author
 cathy labrish posted on Wednesday, April 22, 2015 - 2:50 am
I want to fit a model predicting a distal outcome as follows:
1) one part of the model is a GMM in which latent class also predicts a binary the outcome. Three covariates are regressed both on the values of the latent growth parameters (with regressions allowed to vary across class) and the latent class variable.
2) a second part of the model predicts the distal outcome based on two additional variables which are "unrelated" to the latent class variable (ie. I want to regress the distal outcome on both the latent class variable, and the two additional covariates, but do not want to regress any variable in the GMM on these additional covariates).
I am wondering if the following is the correct specification for this model (i've only provided key snippets):

categorical is outcome;
%overall%
i s| av1-av5;
i s *;
av1-av5;
i on cvar1-cvar3;
s on cvar1-cvar3;
i with s;
c on cvar1-cvar3;
outcome on cvar4 cvar5;
%c#1%
i s| av1-av5;
i s *;
av1-av5;
i on cvar1-cvar3;
s on cvar1-cvar3;
i with s;
%c#2%
i s| av1-av5;
i s *;
av1-av5;
i on cvar1-cvar3;
s on cvar1-cvar3;
i with s;
 Bengt O. Muthen posted on Wednesday, April 22, 2015 - 2:09 pm
Looks alright. The i s | av1-av5 statement needs adjustment (time scores) and having class-varying i s on statements isn't always easy to estimate (too much flexibility).
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: