

A parameterization problem 

Message/Author 

cchien posted on Wednesday, April 21, 2010  9:58 am



Hi. Previously I used a 3factor EFA result in a CFA framework with a survey data set, and everything was fine. While using a 4factor EFA result in a CFA framework, Mplus showed an error message: *** ERROR in MODEL command Variances for categorical outcomes can only be specified using PARAMETERIZATION=THETA with estimators WLS, WLSM, or WLSMV. Variance given for: RSVAR24 I alternatively added the statement PARAMETERIZATION=THETA in my program, but the output was mostly different from the output with 3 factors. My question is, why can the default PARAMETERIZATION=DELTA not be used in 4 factors case? Why are the two outputs of parameterization methods different? Thanks for any answer or information. 


It is not that Delta cannot be used with four factors. It is that you can't mention a variance for a categorical variable. The same thing would happen for three factors. You should remove RSVAR24 from the MODEL command. Or if it is a factor indicator, you should have it as part of a BY statement. If you can't see the problem, send the full output and your license number to support@statmodel.com. 

Back to top 

