I am not an expert in multivariate analysis, son I need your help. I have many covariance matrices related to the items of a score. I have information from many populations. I estimated the eigenvalues for the score in each population using Exploratory Factor Analysis , but I want to test these differences. I have read some ideas from other fields and I know what there are many test to compare matrices, but what I just want to know if the differences obtained in the eigenvalues (and finally in the factorial structure) are significative or not. If It could be run in STATA, It will be great! Thanks...