The University of Oklahoma Heath Services Center
Pseudo-population bootstrap methods for imputed survey data
Item nonresponse in surveys is usually dealt with through single imputation. Treating the imputed values as if they were observed values may lead to serious underestimation of the variance of point estimators. Two pseudo-population bootstrap approaches are used for deriving bootstrap variance estimators: the nonresponse model approach and the imputation model approach. We establish the asymptotic properties of the resulting bootstrap variance estimators for population totals and population quantiles. Results from a simulation study suggest that the proposed methods perform well in terms of relative bias and coverage probability.
This is a Joint work with Professors David Haziza, Christian Leger (University of Montreal) and Zeinab Mashreghi (University of Winnipeg)
Refreshments at 3:45pm in Snedecor 2101.