How well does number of parents explain unique variance in self-reported drug use?