When there is a strong prior reason to believe one thing is better, we don't need too many studies to tell us it is so. If, a-posteriori, we saw that you couldn

When there is a strong prior reason to believe one thing is better, we don't need too many studies to tell us it is so. If, a-posteriori, we saw that you couldn't tell too much difference between college graduates and not by looking at their income, or if you couldn't guess whether someone was EBF as a baby by their health status, you might start wondering if the effect sizes are really so large, despite your prior beliefs.