Newman-Keuls test

n. A post hoc procedure in an analysis of variance in which group means are ranked and a t statistic is used to test the difference between the largest and smallest groups' means, the smallest and second largest, and so on, as long as significant differences are found.