I am getting an error message when performing a CFA multiple group analysis (comparing males and females) using the following syntax:
Factor1 BY var1 var2 var3 var4; Factor2 BY var5 var6 var7; Factor3 BY Factor1 Factor2;
Factor3 represents a "dimensional" factor which comprises factors 1 and 2.
The first two syntax lines run fine but the "Factor3" seems to be the problem - as suggested by the Tech printout. I don't have any problems when running the 3 syntax commands on either the male or female group data individually. But when I run the syntax across both groups (using mgroup)that's when the problem occurs.
Thank you for the help, Dr. Muthen. I don't understand, though, why the second-order factor worked when using data from one group (e.g., males). No error messages was generated. It seemed to have worked with two (not three) first-order factors.
But now the second-order factor with two-first-order factors does not work when using data with both groups (males and females). Could you please explain why the second-order factor having two first-order factors worked in one case but not the other?
You must have had some other restriction in the earlier model. I would have to see both outputs to see what the difference is. A second-order factor model with two first-order factor indicators is not identified.