I have a question regarding what to include in the ascii file from SPSS to be used in Mplus. I've run both an EFA and a CFA to explore and confirm latent constructs, however now that I want to run a path analysis, I'm having some problems. In the EFA and CFA, I included only the variables that were being exmained. I did not include the dependent variables or any control variables in the original EFA and CFA since I didn't want them to be part of the latent consructs. My main question is should I keep these variables in the ascii file even though I don't want them to be used for the EFA and CFA. But, I do want to use them in the path analysis. Any thoughts?
Observed factor indicators in an SEM model can be continuous, censored, binary, ordered categorical (ordinal), counts, or combinations of these variable types. You can use the CENSORED, CATEGORICAL, and COUNT options of the VARIABLE command to specify the scale of dependent variables that are not continuous.