I am trying to figure out how JMP can provide Multivariate tests for homoscedasticity, and univariate tests for homoscedasticity.
So i have moderation of distribution x5 by a firm size x3 and my dependent variables are X19, X20, X21.
JMP does not provide for Box's m test but i can use Barlett. However, using to use both X5 and X3 ni the fit group cannot be done to get to these values for univariate. But i can also get Levene tests here. Can fit group test for Multivariate and univariate for homoscedasticity across groups X5 by X3?