Non-Response in Sample Surveys - A New Approach

Abstract

In performing sample surveys, the investigator often faces the problem of non-response and the possibility of bias as a result of this phenomenon. This article discusses the issue of estimating the population mean in the presence of non-response under simple random sampling scheme using two auxiliary variables. These missing observations may occur on auxiliary, study or both of the variables. A weighted ratiocum-exponential factor-type estimator is suggested and its properties have been studied under large sample approximation. The mean square error of suggested estimator is minimum as compared to selected estimators in the study. The theoretical results are verified through a simulation study