Hello,
I have a question about the boxM test again.
I was reading that to check the equality of correlation matrices I need to transform the raw data to standard scores for each sub-prediod before running the boxm test.
So I use the scale() function for each period and running the boxM test on correlation matrices but I get the same answer no matter if I use the original correlation matrices or the covariance matrices of the respective periods.
For example if I use as input either the correlation either the covariance or the scale(correlation) matrices I get the exact same numerical answer once I pass it ot the boxM function.
Thank you!