An efficient class of calibration ratio estimators of domain mean in survey sampling (Q2006958)

A finite population is divided into \(D\) domains. Stratified sampling for selection of units is adopted and data on the study variable and the related auxiliary variable is collected for the units that lie in a domain of interest. It is desired to estimate the population mean using a calibrated ratio estimator. The calibration weights are chosen such that the chi-square-type loss is minimized subject to the calibration constraint. The bias and estimated variance are derived and compared with some of the conventional ratio estimators in the literature as well as the regression estimator by a simulation study. It is concluded that the proposed class of calibrated ratio estimators fares better.