In his paper 'Model Modification" (Psychometrika, 1989), Dag Sörbum gives a formula for the Modification Index in equation (8). He mentions that a simplified formula is used in LISREL V. I would like to know if the same simplified formula is used in AMOS.
Thank you Michael for your answer. I give more details on my question in the attached file. It seems that AMOS is using the simplified formula for MI.
What do you think?
The answer in the IBM SPSS website reproduces a description of the MI given in the LISREL 6 user's guide (page I.42. The explanation given in the first sentence corresponds exactly to the simplified modification index expressed in term of the fit function FML. But then, the second sentence should be "It can be shown that this index is a lower bound of the expected decrease in chi square if a single constraint is relaxed..." and not "...is equal to...".
What do you think?
Thanks a lot for your help.
Michel Tenenhaus
- Badges
- Science topic
- Similar topics
- Medicine
- Physiology
- Biosignals
- Biomedical Signal Processing
- Statistical Analysis