How Likely Is Simpson's Paradox?

被引:31
作者
Pavlides, Marios G. [1 ]
Perlman, Michael D. [2 ]
机构
[1] Frederick Univ Cyprus, CY-1303 Nicosia, Cyprus
[2] Univ Washington, Dept Stat, Seattle, WA 98195 USA
关键词
Bayes factor; Bayes test; Dirichlet distribution; Multinomial distribution; Simpson's Paradox; Simpson reversal; SUBDIVISIONS; PROPORTION;
D O I
10.1198/tast.2009.09007
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
What proportion of all 2 x 2 x 2 contingency tables exhibit Simpson's Paradox? An exact answer is obtained for large sample sizes and extended to 2 x 2 x l tables by Monte Carlo approximation. Conditional probabilities of the occurrence of Simpson's Paradox are also derived. If the observed cell proportions satisfy a Simpson reversal, the posterior probability that the population parameters satisfy the same reversal is obtained. This Bayesian analysis is applied to the well-known Simpson reversal of the 1995-1997 batting averages of Derek Jeter and David Justice.
引用
收藏
页码:226 / 233
页数:8
相关论文
共 20 条