EFFICIENT RANDOM IMPUTATION FOR MISSING DATA IN COMPLEX SURVEYS
Randy R. Sitter
Simon Fraser University
Abstract
A simple adjusted random imputation method for handling item nonresponse in
complex surveys is presented. This method eliminates the imputation variance
of the estimator of a mean or total, and at the same time preserves the distribution
of item values. Jackknife and bootstrap variance estimators that depend only on
the reported values in the data file are also proposed. It is necessary to identify
the respondent and imputed values in the data file as well as the imputation class.
Simulation results on the performance of the proposed method in estimating a total,
population variance and distribution function are also presented.
(Joint work with J.N.K. Rao and Jiahua Chen)