Message/Author |
|
|
I have a question about how one can best do multigroup analyses when you also have a categorical dependent variable in your structural equation model (and hence use the integration=montecarlo option)? since regular multiple group analyses are not possible with the integration command, I'm considering dividing the data into two groups with the subpopulation option; I first used the useobservations option but I got a warning that the standard errors might be wrong and that I should use the subpopulation option, but is this correct? Or would someone know a better way of dealing with this? |
|
|
I think you can use the KNOWNCLASS option and TYPE=MIXTURE. When all classes are known, it is the same thing as multiple group analysis. |
|
|
Thanks Linda for your response! I tried to perform these analyses but I'm having trouble with my syntax. Do you have any examples on how you should adjust the syntax for a regular structural equation model(with type is complex) to a mixture model? |
|
|
Example 8.8 shows how to use the KNOWNCLASS option. If you have problems, send the output and your license number to support@statmodel.com. |
|
Back to top |