Generalized Almost Unbiased Estimator of Finite Population Variance under Stratified Random Sampling
Isah Muhammad *
Department of Statistics, Binyaminu Usman Polytechnic, Hadejia, Nigeria and Department of Statistics, University of Ilorin, Ilorin, Nigeria.
Gafar Matanmi Oyeyemi
Department of Statistics, University of Ilorin, Ilorin, Nigeria.
*Author to whom correspondence should be addressed.
Abstract
This study developed a generalized almost unbiased estimator for obtaining finite population variance under stratified random sampling. The generalized estimator was developed using the approach of almost unbiased. This approach uses filtration parameters in reducing the bias of estimators. The theoretical properties of the generalized estimators were derived along with the expressions of the filtration parameters. The efficiency conditions of the proposed estimator over some existing population variance estimators were obtained theoretically. The performance of the proposed estimator was evaluated empirically using simulated and real-life data sets. The generalized estimator performed better with minimum values of bias, mean square error, and higher percentage relative efficiency. Therefore, the generalized almost unbiased estimator can be utilized to provide better variance estimates for various phenomena such as inflation variation over years, exchange rate variation over years and standard of living variation over for better policymaking.
Keywords: Estimator, variance, unbiased, mean square error, efficiency