Estimation of the best linear unbiased predictor for the mean with unequal sample sizes

被引:0
|
作者
Rao, Poduri S. R. S. [1 ]
机构
[1] Univ Rochester, Stat Program, Rochester, NY 14627 USA
关键词
Stein-type estimator; Predictors for the means; Bias and mean square error; Fixed values of the means; Performance of the EBLUP; Hospital expenses;
D O I
10.1016/j.stamet.2012.02.001
中图分类号
O21 [概率论与数理统计]; C8 [统计学];
学科分类号
020208 ; 070103 ; 0714 ;
摘要
When the samples selected from k normal populations are of unequal sizes, we consider the empirical best linear unbiased predictor, EBLUP, for the mean of each population. For fixed values of the means of these populations, conditions for the Mean Square Error (MSE) of the EBLUP to be smaller than the variance of the sample mean and, at the same time, for its absolute bias to be smaller than a specified fraction of the square root of its MSE are obtained. Preference of the EBLUP over the sample mean is examined for the estimation of the averages of the daily hospital expenses of the Standard Metropolitan Statistical Areas (SMSAs) of twenty states in the US. (C) 2012 Elsevier B.V. All rights reserved.
引用
收藏
页码:515 / 519
页数:5
相关论文
共 50 条