Health scientists increasingly use data from unequal-probability survey designs, either public-use data
from national surveys such as NHANES and NLSY, or their own surveys, or from subsamples of existing
cohorts or databases. Correct analysis of survey data requires appropriate software and an
understanding of basic survey concepts, but is otherwise just like any data analysis. In this module we
will cover the concepts of weights, clusters, and strata, and how to use the R survey package to conduct
analyses.