We have learned that in order to compare two independent groups on a quantitative outcome of interest, we need to perform an independent samples t test. When the number of groups exceeds 2 we need to perform analysis of variance (ANOVA). Think carefully about the relationship between ANOVA and independent samples t test. If there are three groups (say A, B, and C) then what is wrong with performing three pairwise mean comparisons (A versus B, A versus C, and B versus C) using independent samples t test instead of performing an ANOVA that simultaneously compares the three groups?