. . "The answer depends on how large of an effect size (in these examples difference between means), how much variability, the specific statistical test you are doing and other factors like whether it is a one or two-tailed tests and correlations between your groups." . .