On the Adjustment of Non-Response through Imputation for Estimating Current Mean in Repeated Surveys

Abstract

In this paper we have proposed an imputation method based on a family of factor-type estimator to deal with the problem of non-response assuming that the target population has been sampled at two different occasions. The aim is to estimate the current population mean on the basis of matching the sample from the previous occasion and on the basis of fresh sample selected at the current occasion. It has been assumed that the non-response is exhibited by the population at both the occasions and, therefore, the imputation of missing values is required in both the samples, namely, matched sample and fresh sample. Accordingly, a combined point estimator has been suggested after imputation which generates a one-parameter family of estimators. The properties of the estimator have been investigated and the replacement policy has been discussed. Finally, the comparison of the proposed class has been made with another estimator for their performances.