All analyses incorporate sampling weights to adjust for oversampling of some demographic groups, and all standard errors are adjusted to correct for dependencies in the data caused by the complex survey design (that is, unequal probability of selection into the sample through clustering by geographically-defined stratum and primary sampling unit).
