LPA with composite index? PreviousNext
Mplus Discussion > Latent Variable Mixture Modeling >
 Mplus fan one posted on Saturday, May 10, 2008 - 5:31 pm

I wonder if there is any merit--or cause for concern?--in conducting a latent-profile analysis (LPA) using a composite score derived by principal-components analysis (PCA) of several continuous measures.

I've looked at some refs and was surprised to find even a handful of papers and dissertations that describe such analysis.

Hope kind readers may give some insight!
 Bengt O. Muthen posted on Sunday, May 11, 2008 - 10:37 am
If you mean that you do an LPA on a set of variables that are PCA scores I would be concerned because such scores have been created as uncorrelated and LPA is a model for correlated variables.
 Mplus fan one posted on Sunday, May 11, 2008 - 12:32 pm
Hello, Prof. Muthén--thank you for 'connecting' those fundamental aspects of these paradigms!
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