Use of imputation methods at current occasion in two-occasion rotation patterns

Abstract

Among modern strategies used to cope with missing data, a major problems faced by survey statisticians, imputation is one of the most common. This technique consists of replacing missing values with their possible estimates so that the resulting data set is complete and can be analyzed using standard methods. The objective of the present investigation is to study the use of imputation techniques in dealing with non-response at current occasion in two-occasion successive (rotation) sampling. Two difference-type estimators are proposed for estimating the population mean at current occasion. Detailed behavior of the proposed estimators is studied and optimum replacement policies are discussed. To study effectiveness of the suggested imputation techniques, performances of the proposed estimators are compared in two different situations: with and without non-response. Empirical studies are carried out to demonstrate the performance of the proposed estimators.