Confounding Variables In Statistics