Double sampling regression estimator of population mean subsequent to preliminary test of significance in finite population survey sampling

摘要

A preliminary test double sampling regression estimator (PTDSRE) in finite survey sampling has been developed when the prior value of the correlation coefficient between a study variable and an auxiliary variable is available. The PTDSRE has been found to be the unbiased estimator of population mean. Its variance has been derived. Numerical findings showed that PTDSRE has outperformed the usual double sampling difference and regression estimators.