All Times ET
Program is Subject to Change
Variance Estimation for Nonprobability Surveys (308336)*Jill Dever, RTI
Variance estimation for probability surveys relies on an established theoretical foundation. By contrast, there is currently no universally excepted variance estimation approaches (or theory) for nonprobability surveys. In the second half of this lecture, we summarize weighting methodologies used with nonprobability survey data to produce population estimates. Building on this information and details from the first half of the lecture, we review and compare a few approaches for variance estimation such as quasi-randomization, prediction theory, and model based.