Data from surveys should be weighted using expansion weights to avoid
biased estimates for the inference population. Calculation of expansio
n weights is a stepwise procedure that can be thought of as allowing r
espondents to represent all eligibles from the population. Design will
have an impact on the variance of the estimates, Furthermore, the imp
act may be to inflate or deflate the variance and is often variable-de
pendent. Accounting for the design in estimating the variance will req
uire specialized analysis procedures and often specialized software.