Getting posterior probabilities for e... PreviousNext
Mplus Discussion > Latent Variable Mixture Modeling >
Message/Author
 Brian Segal posted on Wednesday, March 19, 2014 - 1:30 pm
I am working on an LTA following example 8.13 in the user's guide. Is it possible to output the posterior class membership probabilities for each subject at each time point?

For example, suppose data were collected on 1,000 subjects over 3 time points, and that I'm modeling the data with 2 latent classes. Then I would like to output the posterior probabilities of each subject belonging to each latent class at each time point, for a total of 1,000*3*2=6,000 posterior probabilities.

Thank you for your help.
 Bengt O. Muthen posted on Wednesday, March 19, 2014 - 4:21 pm
I think you get that in the Savedata command, saying

Save = cprob;

and specifying File = name;
 Brian Segal posted on Thursday, March 20, 2014 - 3:20 pm
Fantastic, thank you again for your help. It looks like cprob outputs the posterior probabilities of each latent class pattern as well as the most likely class at each time point. Is it possible to also output the posterior probabilities at each time point? If not, I'll just marginalize the latent class pattern posteriors to get posteriors for individual time points.

Thanks.
 Linda K. Muthen posted on Friday, March 21, 2014 - 9:46 am
You would need to do this yourself.
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: