Hi everyone, I'm running into a problem with my path analysis in Mplus working off summary data. I'm looking at two different outcome variables, and I'm testing models where I just swap out one outcome for the other while keeping the rest of the model the same. My problem is that I get perfect fit indices with out outcome but not for the other outcome (identification doesn't seem to be an issue because the # of variables is the same).
Basically, I'm looking at 6 variables (A through F) in two different models in which I keep 4 of the variables and swap out one of the last 2, keeping the models the same except for the exchanged variables. Here is the structure:
A B C all interrelated Paths from A B C to D Paths from A B D to E/F (E and F are different variables, models are tested with either E or F, but not both).
For outcome E, I get perfect fit indices while I don't for outcome F. Why is this happening, or, what are some things that can contribute to this happening?