Testing Fisher, Neyman, Pearson, and Bayes

被引：82

作者：

Christensen, R ^{[1
]}

机构：

[1] Univ New Mexico, Dept Math & Stat, Albuquerque, NM 87131 USA

来源：

AMERICAN STATISTICIAN | 2005年 / 59卷 / 02期

关键词：

confidence; Lindley's paradox; most powerful test; p values; significance tests;

D O I：

10.1198/000313005X20871

中图分类号：

O21 [概率论与数理统计]; C8 [统计学];

学科分类号：

020208 [统计学]; 070103 [概率论与数理统计]; 0714 [统计学];

摘要：

This article presents a simple example that illustrates the key differences and similarities between the Fisherian, Neyman-Pearson, and Bayesian approaches to testing. Implications for more complex situations are also discussed.

引用

页码：121 / 126

页数：6

共 13 条

[1]

Teaching Bayes' rule: A data-oriented approach [J].

Albert, J .

AMERICAN STATISTICIAN, 1997, 51 (03) :247-253

[2]

[Anonymous], 1971, The Design of Experiments

[3]

Berger J.O., 1984, LIKELIHOOD PRINCIPLE, V6

[4]

Could Fisher, Jeffreys and Neyman have agreed on testing? [J].

Berger, JO .

STATISTICAL SCIENCE, 2003, 18 (01) :1-12

[5]

Bayesian statistics and the efficiency and ethics of clinical trials [J].

Berry, DA .

STATISTICAL SCIENCE, 2004, 19 (01) :175-187

[6]

Teaching elementary Bayesian statistics with real applications in science [J].

Berry, DA .

AMERICAN STATISTICIAN, 1997, 51 (03) :241-246

[7]

SAMPLING AND BAYES INFERENCE IN SCIENTIFIC MODELING AND ROBUSTNESS [J].

BOX, GEP .

JOURNAL OF THE ROYAL STATISTICAL SOCIETY SERIES A-STATISTICS IN SOCIETY, 1980, 143 :383-430

[8]

Significantly insignificant F tests [J].

Christensen, R .

AMERICAN STATISTICIAN, 2003, 57 (01) :27-32

[9]

Ferguson TS., 1967, Mathematical statistics: A decision theoretic approach

[10]

Fisher R.A., 1973, Statistical Methods and Scientific Inference, V3rd

← 1 2 →