結果 : what is used to reduce the effects of confounding variables in experiments