Tom Krenzke and Jianzhu Li

Estimates in published tabular results from surveys sometimes display, and should display, an associated standard error (measure of variation due to sampling). Using the published point estimates in a subsequent analysis should bring with it the associated variation that propagates through so the user knows how much to trust the subsequent analysis results. Failure to do so can mislead researchers and policymakers due to the inaccurate precision estimates in the results.

Some content is only viewable by ASA Members. Please login or become an ASA member to gain access.