Tukey Method

The Tukey method is a multiple-comparison procedure designed specifically for all-pairs comparisons of treatment means after ANOVA.

It uses the Studentized range distribution rather than a $t$ critical value. Compared with Bonferroni, Tukey is often less conservative for all-pairs inference.

R

fit <- aov(y ~ treatment, data = dat)
TukeyHSD(fit)