Improved estimators for mean of population using ancillary variable under linear systematic sampling

Abstract: The information based on ancillary variable is essential for estimation of the mean of population in surveys related to the agriculture, medical and forest. Numerous ancillary variables information, including the correlation coefficient, coefficient of variation and kurtosis etc. might affect the estimators’ performance. Many studies have been conducted using such information to estimate the mean of population under linear systematic sampling (LSS). There is a dearth of research on estimating the mean of population in LSS using the known coefficient of variation and intra-class correlation coefficient of the study variable. Thus, in the context of LSS, improved estimators for the mean of population have been suggested, utilizing the known coefficient of variation and intra-class correlation related to the study variable. These novel estimators’ properties are examined, and the conditions under which they achieve better than their counterparts are determined. Furthermore, empirical investigation is done to evaluate the impact of the constructed estimators in comparison to the relevant estimators.

Keywords: Ancillary variable, Bias, MSE, Coefficient of Variation.

Mathematics Subject Classification  62D05 · 62G05 · 62H12