LAUSR

A nalysis of variance (ANOVA) is a statistical technique for comparing the means of samples. In its simplest form, one-way ANOVA, it is typically used to compare the means of three or more samples as oneand two-sample t tests, respectively, are generally used to compare a single sample mean against a standard or two samples’ means against each other. With three samples, complexity arises comparing the means of groups A, B, and C using repeated two-sample t tests. If A is not different from B and B is not different from C but A is different from C at one chance in 20 (p < 0.05), what does it mean that three comparisons were made and not one? Neither of the easy answers, ignoring the existence of B or multiplying the A versus C probability by three because three measures were made, is either satisfying or accurate.