Small studies can give misleading results because of random chance - like flipping a coin a few times might not show the real pattern.