Categorical and Count outcomes PreviousNext
Mplus Discussion > Structural Equation Modeling >
Message/Author
 DW posted on Monday, February 20, 2012 - 12:24 am
Hello,

I apologize because I know I am asking a number of questions. I am hoping to run an SEM in which I have 3 outcome variables. One outcome variable is an observed dichotomous outcome and the other two are observed count variables (and should probably be considered zero-inflated). I wonder if it is possible to run an SEM with these different types of outcomes or is it not possible due to different estimators being required? Currently, I have been running them separately and have had no problem running the SEM with just the dichotomous outcome using WLSMV.

However, I cannot seem to get an SEM to run with the just the count outcome variables. I get the error message "THE MODEL ESTIMATION DID NOT TERMINATE NORMALLY DUE TO A NON-ZERO
DERIVATIVE OF THE OBSERVED-DATA LOGLIKELIHOOD.
THE MCONVERGENCE CRITERION OF THE EM ALGORITHM IS NOT FULFILLED.
CHECK YOUR STARTING VALUES OR INCREASE THE NUMBER OF MITERATIONS.
ESTIMATES CANNOT BE TRUSTED. THE LOGLIKELIHOOD DERIVATIVE
FOR PARAMETER 93 IS -0.69959041D-01." I have increased the number of iterations and that does not seem to help. The model runs fine if I treat the variables as just a continuous variable. I am hoping you could please provide some guidance.

Thank you so much for your help!
 Linda K. Muthen posted on Monday, February 20, 2012 - 7:58 pm
You can analyze a combination of categorical and count variables using maximum likelihood estimation.

Please send the output and your license number to support@statmodel.com.
Back to top
Add Your Message Here
Post:
Username: Posting Information:
This is a private posting area. Only registered users and moderators may post messages here.
Password:
Options: Enable HTML code in message
Automatically activate URLs in message
Action: