We now know how to compare two groups on a single outcome, like males and females in terms of blood pressure, or Diet A vs. Diet B on hours of sleep. But we often want to compare an outcome across many groups. We might want to look at the 50 US states in terms of their income levels: can we say which states have statistically significantly higher or lower incomes? Or we might want to look at the effects of 6 different types of exercise (biking, hiking, tennis, weightlifting, walking and swimming) on glucose levels, and ask if any of the forms of exercise have statistically significantly better outcomes.