Message/Author |
|
|
How do I use the modification indices for BY and WITH statements to improve my model? Thank you!!! |
|
|
The modification index is the value that chi-square will drop if you free the corresponding parameter. You should not free a parameter unless it makes sense to do so and you can rationalize doing it. If you have not already done an EFA on your data, I would suggest this as a first step. It is easier to correct problems using EFA results than to try to fix an ill-fitting CFA model. |
|
|
How is it possible that the expected change in the chi-square statistic would exceed the value of the chi-square that resulted from a particular analysis? For example: H0 gives chi-square on 2DF = 9.81. One MI value is reported at 108.22. It seems that something is amiss, or? |
|
|
Are you looking at the M.I. column or the E.P.C. column. If the M.I. column, please send your output and license number to support@statmodel.com. The E.P.C. column is the expected parameter change. |
|
Back to top |