Model Fit for 1 Categorical & 2 Conti... PreviousNext
Mplus Discussion > Categorical Data Modeling >
 Deanna Ibrahim posted on Friday, November 15, 2019 - 11:35 am

I am using SEM to predict pathways from my IV to 3 different DVs (1 categorical, 2 continuous). I had originally estimated the model using ML (without specifying that one of my outcomes was categorical), and the model fit was sufficient. However, when I specified the categorical outcome (via CATEGORICAL IS) and used WLSMV estimation, the model fit got much worse, and the path estimates for my continuous variables were quite different. I'm wondering if you could let me know 1) why the model fit would look much worse when estimating with WLSMV, and 2) whether it would be appropriate to use MLR estimation, since I do have only one categorical outcome.

Thank you!
 Bengt O. Muthen posted on Friday, November 15, 2019 - 2:02 pm
Perhaps your sample size is low so that the WLSMV chi-square doesn't work well. We need to see the 2 outputs you refer to - send to Support along with your license number.
 Deanna Ibrahim posted on Saturday, November 16, 2019 - 7:15 am
I will do that- thank you.
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