In statistics, a method for comparing treatment group means after The analysis of variance (ANOVA) null hypothesis that the expected means of all treatment groups under study are equal, has been rejected using the ANOVA F-test. If the F-test fails to reject the null hypothesis this procedure should not be used.