Power transformation design-based calibration scheme of robust measures under stratified sampling

Abstract

In this article, two modified design-based calibration scheme-type estimators are formulated. The suggested estimators were proposed under robust measure of stratified sampling as information of auxiliary variables is sensitive to outliers on data. The technique such as coefficient of variation, Gini mean diference, Downton method, Probability waited moment, Median of the class, Hodge-Lehmann, first quarter, first decile, second decile, third decile were handle presence of outliers. The metric means square error (MSE) of the suggested calibrated scheme-type estimators are deduced up to first order of approximation by new Tailor series approach. The conditions at which the proposed calibrated scheme-type estimators outperformed existing estimators counterparts are studied numerically using simulated data under two populations, which were generated using Log-normal and Weibull distributions respectively with R scripts. The proposed calibrated scheme-type estimators efficiency in a rows is kept continue up to last population (Weibull) without any exception.