Variance Estimation for the Population Total Estimator in the Presence of Nonresponse

摘要

This paper considers variance estimation for the population total estimator focusing solely on the unit nonresponse and when sampling fraction is negligible. Unit nonresponse may occur in a sample survey and it was handled by the weighting adjustment technique. Consequently, the population total estimator leads to the nonlinear estimator. Hence, its variance can be derived by using the Taylor linearization procedure. A method to estimate the variance of population total estimator is proposed under simple random sampling. Our method neither requires known or estimated response probability under a reverse framework and when sampling fraction is negligible. Finally, the proposed estimator is compared with Arnab and Singh's estimator using mean square error.