In lots of human studies, you can’t just stop at an arbitrary number of participants because you’ve counterbalanced manipulations to decorrelate potential confounders (e.g., which color stimulus is paired with reward, the order of trials).