Stratified calibration estimator of population mean using multi-auxiliary information

Calibration estimator uses calibrated weights that minimize a given distance measure to the design weights while satisfying a set of constraints based on the known auxiliary information such as population total (or mean). In this paper, we propose an improved multivariate calibration estimator of population mean in stratified sampling design. The problem of determining the calibrated weights is solved using Lagrange multiplier technique. The proposed multivariate calibration estimator of population mean is derived in the form of linear regression estimator (LREG). A numerical example is presented to illustrate the application and computational details of the proposed estimator.