In this study, we compared two cohorts of women and their infants, one that took Methamphetamine during pregnancy and one that did not. It was desirable for the unexposed group to be representative of the hospital population. Therefore, we chose to samples hospital deliveries systematically by including the next consecutive birth after the Methamphetamine-abusing mother/infants pair. When sampling is systematic with respect to a variable not correlated with study parameters, one would expect the sampling and to have the same variance.