A Permutation Test for Multivariate Matched-Pair Analyses: Comparisons with Hotelling's Multivariate Matched-Pair T ² Test

Abstract

A permutation alternative for Hotelling's multivariate matched-pair T 2 test is introduced. The permutation test allows for analyses when the number of subjects is less than or equal to the number of measurements, which is not possible with Hotelling's multivariate matched-pair T 2 test. For the data analyzed the permutation test is shown to provide improved discrimination over Hotelling's multivariate matched-pair T 2 test.