AN EXACT UNCONDITIONAL TEST FOR THE 2X2 COMPARATIVE TRIAL
AN EXACT UNCONDITIONAL TEST FOR THE 2X2 COMPARATIVE TRIAL
复制标题
DOI：
10.1037/0033-2909.99.1.129
复制
发表时间：
1986-01-01
影响因子：
22.4
通讯作者：
HABER, M
中科院分区：
文献类型：
--
作者：
HABER, M
基于共同Z统计量的无条件分布，提出了一种单侧精确检验，用于假设在2 × 10 - 5的情况下成功概率相等。2项比较试验列联表。大样本（正态）检验的第1类错误概率可能超过标称显著性水平的两倍，这一事实证明需要精确检验。功效比较表明，新检验的性能明显优于Fisher精确检验。在某些情况下，精确Z检验甚至比随机条件检验更强大，随机条件检验在所有无偏检验中一致是最强大的。
A one-sided exact test, based on the unconditional distribution of the common Z statistic, is proposed for the hypothesis of equal success probabilities in a 2 .times. 2 comparative trials contingency table. The need for an exact test is justified by the fact that the Type 1 error probabilities of the large-sample (normal) test may turn out to be more than twice the nominal significance level. Power comparisons reveal that the new test performs considerably better than Fisher''s exact test. In some cases, the exact Z test is even more powerful than the randomized conditional test, which is uniformly most powerful among all the unbiased tests.