Latent variable regressions in LTA PreviousNext
Mplus Discussion > Latent Variable Mixture Modeling >
Message/Author
 Mark Burton posted on Saturday, November 19, 2016 - 2:07 pm
Hello,

I am conducting a LTA with 4 classes at two time points. The output provides coefficients for the regression of each level of C2 on C1 compared to one of the levels as a reference class. I would like to explore the likelihood of transitioning to one C2 class vs all other classes within the latent variable. Essentially, I would like to make my reference class "not transitioning to a specific class" compared to "transitioning to that class". Is there a way to run posthoc analyses where I create a dichotomous C2 variable to explore this?

Thank you.
 Bengt O. Muthen posted on Monday, November 21, 2016 - 4:54 pm
Use Model Constraint to express any transition probability you like using the labels of the logit parameters. See also our web note 13.
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: