It is assumed that the -dimensional data in the groups have a multivariate normal distribution: distributed with expected value vectors and covariance matrices ( ).
The hypothesis to be tested that all covariance matrices are the same, so
vs. there are min. a couple and with .
The test variable for the test is the so-called M from Box,
in which
serves as a correction. The covariance matrix is from the observations to the group , estimated include
and the pooled, i.e. mean, covariance matrix
If the test variable is sufficiently large, it is approximately distributed in chi-squares with degrees of freedom . If they are very different from overall , the test quantity value becomes high. is therefore rejected at the level of significance if M is greater than the - quantile of the chi-square distribution with degrees of freedom.