Jeroen Ooms posted on Thursday, August 28, 2008 - 4:29 am
In a little MC study I have evaluated the power off the common model of fit tests (chisq, TLI, CFI, RMSEA), on multigroup data when one of the factor loadings is not homogeneous across the groups. I have found that their power is often too small to detect these variations.
Now I would like to evaluate if the modification indices give a better indication. I have tried the OUTPUT: MODINDICES; option, however this does not seem to work in a MC study. Is there any way to output modification index summary statistics of the parameters in a MC analysis?
The only way to do this would be to save all of the data sets and run an external Monte Carlo. See Example 11.6. You would then have to run each data set separately to get the outputs and modification indices.