LCA / LTA observed variables PreviousNext
Mplus Discussion > Latent Variable Mixture Modeling >
 Johan Claas posted on Thursday, June 05, 2014 - 5:01 am
Dear community,

for my thesis project I am just getting started with latent transition models. I have a fairly basic question, which I have not found explicitly answered yet on the internet.

When I am fitting an LTA model there are different ways of incorporating observable variables. I would like to know what the difference is between using a variable as

1. (additional) manifest indicator directly in the LTA / LCA
2. covariate
3. grouping variable in multigroup LTA / LCA.

I understand the modelling results (e.g. parameters estimated) will differ in each case but what is, qualitatively speaking, the difference?

 Bengt O. Muthen posted on Thursday, June 05, 2014 - 9:45 pm
1 vs 2. Think of LCA. If you have say 2 covariates you don't specify them as uncorrelated given the latent class variable, but you do if they are indicators. With 1 covariate there is no difference.

A grouping variable is like a covariate; in fact, in LCA with categorical outcomes you can do everything you can do with groups (Knownclass) also with covariates.
 Johan Claas posted on Friday, June 06, 2014 - 2:27 pm
Thank you, that makes it a lot clearer.
Back to top
Add Your Message Here
Username: Posting Information:
This is a private posting area. Only registered users and moderators may post messages here.
Options: Enable HTML code in message
Automatically activate URLs in message