Calibration Estimation In Survey Sampling

Abstract

The current investigation is an attempt to study the importance of calibration methodology in survey sampling. The major purpose of the study is to estimate the parametric quantities such as nite population mean, population total and population variance in three di erent sampling schemes by using calibration estimation. We developed di erent calibration estimators with the help of optimal calibrated weights produced by minimizing a distance function while satisfying a set of calibration constraints based on use of the known auxiliary information. E ciency comparison is done numerically with the help of simulation study and real data sets. In Chapter 2, estimation of mean is done by using the dual use of the auxiliary information in strati ed sampling. In Chapter 3, a nite population mean is estimated by minimizing a variance function by using the multi-auxiliary variables in strati ed sampling. In Chapter 4, non-linear calibration constraints of the auxiliary variable are introduced for estimating the nite population mean in simple random sampling, probability proportional to size sampling and strati ed sampling. In Chapters 5 and 6, a novel distance function is developed to estimate the nite population total and population mean in probability proportional to size sampling and strati ed sampling. In Chapter 7, an innovative unbiased estimator of nite population variance is proposed in strati ed sampling and also a new estimator of population variance is suggested by using calibration estimation.