Controlling for Family Variables in Abuse Effects Research