A Regression-Type Estimator Based on Preliminary Test of Significance
本文提出一种基于显著性预检验的回归型估计量，在差分估计量和回归估计量之间进行选择，并评估其相对于其他回归型估计量的效率。
Abstract If data on an auxiliary variable X correlated with the variable Y under study are available, regression-type estimators are often used to estimate the population mean μ y . An estimator based on a preliminary test of significance that chooses between the difference estimator and the regression estimator has been proposed. This article investigates the efficiency of the proposed regression-type estimator with respect to other regression-type estimators. Key Words: Preliminary testRegression-type estimatorEfficiency