CFA with dichotomous indicators PreviousNext
Mplus Discussion > Confirmatory Factor Analysis >
 Rich Mohn posted on Monday, February 24, 2020 - 2:04 pm
So I'm running my first CFA with dichotomous indicators (yes/no as 1/0) and am using the 'categorical are' syntax, and it's working fine . . . however, I've also run it with estimator=mlm, and that also produces quite similar results. There is a preceding EFA that was run in SPSS, which of course can't account for either of those options.

Just looking for some guidance on which is more appropriate for the CFA.
 Bengt O. Muthen posted on Monday, February 24, 2020 - 5:14 pm
You should use Categorical approach.
 Rich Mohn posted on Friday, February 28, 2020 - 1:40 pm
Thanks . . . the first order CFA looks great, and I am trying a second order model and it won't converge with 10000 iterations . . . is there anything special I need to specify since the first order latent variables are built on categorical indicators?

Or is it that the second order model just isn't a good model : ).
 Bengt O. Muthen posted on Saturday, February 29, 2020 - 5:53 am
No, nothing special for categorical and a second-order model.
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