Jackknife empirical likelihood confidence interval for the Gini index

被引:16
作者
Wang, Dongliang [1 ]
Zhao, Yichuan [2 ]
Gilmore, Dirk W. [3 ]
机构
[1] SUNY Upstate Med Univ, Dept Publ Hlth & Prevent Med, 750 East Adams St, Syracuse, NY 13210 USA
[2] Georgia State Univ, Dept Math & Stat, 30 Pryor St, Atlanta, GA 30303 USA
[3] Lifeline Capital Management LLC, 1979 Lakeside Pkwy, Atlanta, GA 30084 USA
关键词
Confidence interval; Lorenz curve; Total time on test; Wilks' theorem; STANDARD ERROR; EXPONENTIAL-DISTRIBUTION; VARIANCE ESTIMATORS; CAUTIONARY NOTE; LORENZ CURVE; INEQUALITY; INFERENCE; COEFFICIENT;
D O I
10.1016/j.spl.2015.09.026
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
Jackknife empirical likelihood for the Gini index is derived. Adjusted jackknife empirical likelihood and bootstrap calibration are further investigated. The resulting interval estimators are comparable to existing empirical likelihood methods in terms of coverage accuracy, but yield much shorter intervals. (C) 2015 Elsevier B.V. All rights reserved.
引用
收藏
页码:289 / 295
页数:7
相关论文
共 29 条