Factor Mixture Model Variations PreviousNext
Mplus Discussion > Latent Variable Mixture Modeling >
 J.D. Haltigan posted on Friday, May 19, 2017 - 10:35 pm
Quick question for the most basic factor mixture model (i.e., FMM-1 in Clark et al., 2013). I notice that the syntax for this example differs slightly from 7.17 in the UG. Namely, in the former, variance of fac explicitly set to zero and the analysis requires numerical integration. I ran them both ways and BIC et al. are identical but factor means depart from one another a bit between the two ways of setting it up. Would this be expected even tho model is essentially the same (i.e., most basic LCFA)?
 Linda K. Muthen posted on Saturday, May 20, 2017 - 6:44 am
Please send the two outputs and your license number to support@statmodel.com.
 J.D. Haltigan posted on Saturday, May 20, 2017 - 12:45 pm
Check that, upon further inspection it was just (presentation of) class ordering that was shifting between the runs (i.e., fac mean 1 in lclass 1 was now fac mean 3 in lclass 3; signs also shifted but that is irrelevant as the factor meaning, loadings, et al., all identical down to the 2nd decimal for all values). Apologies for the false alarm bell.
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