I am trying to compare to structural equation models in Mplus (one indirect effects model and one direct effects model). I have done this before in AMOS, but based on the reviewers comments we are now trying to use Mplus for our analyses. However, in AMOS one model had a good fit and the other fitted bad to the data, while in Mplus they both fit well to the data. Could this have something to do with the fit of the measurement model (so the BY statements) fitting well to the data or do you have any idea how this is possible?
Please send the two outputs and your license number to firstname.lastname@example.org. If you are fitting the same model to the same data using the same estimator, you will get the same results in Mplus and Amos.