Multivariate analysis of variance

Summary

Multivariate analysis of variance, or MANOVA, tests whether several groups have the same mean vector when each observation consists of several measurements. It extends the analysis of variance from one response variable to a vector of responses. It compares a between-group and a within-group matrix of sums of squares and cross-products. Common test statistics are Wilks' lambda, Pillai's trace, the Hotelling–Lawley trace and Roy's largest root.