Confounding Variables And Internal Validity In Research