Class membership as a predictor - alg... PreviousNext
Mplus Discussion > Latent Variable Mixture Modeling >
Message/Author
 Miriam Markus posted on Tuesday, September 12, 2017 - 6:32 am
Good afternoon,

I did a latent profile analysis and saved both the class membership and the class probabilities in a new file.
My aim is to develop a diagnostic tool.
Therefore I want to use my results from the LPA to diagnose new subjects by using the same questionnaire I used for the sample I did the LPA with. I want to allocate these new subjects to the most likely latent class, based on their questionnaire, but without using Mplus again.

Is there any possibility to get an algorithm (based on the LPA I did) to estimate the class membership probabilities for any new sample?
 Bengt O. Muthen posted on Tuesday, September 12, 2017 - 6:28 pm
It isn't a simple formula. See e.g. equations in the Muthen-Shedden article:

Muthén, B. & Shedden, K. (1999). Finite mixture modeling with mixture outcomes using the EM algorithm. Biometrics, 55, 463-469.
download paper contact first author show abstract
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: