I've completed some longitudinal measurement invariance in mplus, and was recently asked to calculate effect sizes of loadings/means. I assume this means looking to see the effect size of change in loadings/means of items across 2 timepoints, but I'm not seeing this well described in the literature. Is it as simple as taking cohen's d for each item by calculating: (difference in item i loadings/means across 2 timepoints) divided by (pooled SD of item i across 2 timepoints)?
Thanks.