Regression-cum-ratio mean imputation class of estimators using non-conventional robust measures

Abstract

Different imputation strategies have been developed by several authors to take care of missing observations during analyses. Nevertheless, the estimators involved in some of these schemes depend on known parameters of the auxiliary variable which outliers  can easily influence. In this study, a new class of ratio-type imputation methods that utilize parameters that are free from outliers has been presented. The estimators of the schemes were obtained and their MSEs were derived up to first-order approximation using the Taylor series approach. Also, conditions for which the new estimators are more efficient than others considered in the study were also established. Numerical examples were conducted and the results revealed that the proposed class of estimators is more efficient.