cancel
Showing results for 
Show  only  | Search instead for 
Did you mean: 
Browse apps to extend the software in the new JMP Marketplace
Choose Language Hide Translation Bar
Poparad
Level II

Box's M Test for homogeneity of covariance?

Hi, I'm working on a homework assignment for a statistics class and I'm tasked with finding whether the data set violated the assumption of homogeneity of covariance for a repeated measures split plot setup. I'm using JMP Pro because it's free through my university, however, the lesson material is structured for using SPSS (which is highly not free). I've been searching high and low for how to do Box's M test in JMP and so far only found one post suggesting that it's not possible. Is it possible, and if so, how would I go about running it? Thanks!

3 REPLIES 3

Re: Box's M Test for homogeneity of covariance?

This test is unfortunately not available in JMP. See this Wikipedia entry about the test. You might be able to use other tests for equal variance. Is your analysis of a single response Y and a single categorical X? If so, try Analyze > Fit Y by X. Enter the data columns in the Y and X roles, and then click OK. Click the red triangle and select Unequal Variance. Box's M test is not among the results but there are several other tests for this hypothesis.

Poparad
Level II

Re: Box's M Test for homogeneity of covariance?

The format for the data had 3 different methods of treatment for the between-subjects variable and a response time measured at four different time periods for the within-subjects variable. So I guess that would make it a single categorical X but a multiple response Y? What are some alternative tests that JMP is able to do? Thanks!

Re: Box's M Test for homogeneity of covariance?

Please see this JMP Technical Note first. It might give you some ideas before returning with your question.