Jack Noone posted on Thursday, October 04, 2007 - 6:34 pm
I have 6 observed categorical variables and in CFA they form two factors (3 to a factor - factors are not related) and the model fits well. I want to use these two factors as dependent variables in a SEM model but i am not sure if this is possible. The model is not identified if i test for second order factors and these factors are only just identified if i run the CFA's seperately for each factor. Do i have a problem? It would be nice to have more indicators but sadly that is not possible. I've just started using mplus and SEM so forgive me if this question seems odd.