Topic: Title: Calibration estimator with new distance functions Authors:  M G M Khan - The University of the South Pacific (Fiji) [presenting]
Abstract: Calibration is a method of adjusting the original design weights that increase the precision of the estimates of a characteristic incorporating the known population parameters of auxiliary variables. In order to minimize a given distance measure, the calibration weights are chosen satisfying the constraints related to the available auxiliary information. In survey sampling, many authors have defined calibration estimators by using different constraints and a chi-square type distance function. For stratified random sampling design, we review the calibration approach and propose some new calibration estimators of population mean using new distance functions by varying calibration constraints. A numerical study was carried out to compare the performance of the proposed estimators with the existing calibration estimators. It revealed that the proposed calibration estimators are useful in increasing the precision of the estimates.