Matsumoto K
Toxicological Laboratories, Japan Tobacco Co., LTD, Kanagawa, Japan.
Nihon Yakurigaku Zasshi. 1997 Dec;110(6):341-6. doi: 10.1254/fpj.110.341.
The cumulative chi-squared statistic has been proposed for testing against ordered alternatives in various statistical models. As usual statistical tests of ordered column categorical data, the chi 2 test, Fisher's exact test and Wilcoxon test are used. Pharmacological studies often are performed by multiple dosing. Data obtained from these studies are called ordered categorical data. The cumulative chi-squared statistic, which has been proposed by Hirotsu and Shibuya for testing against ordered alternatives in various statistical models, is little used in spite of its good applicability in the field of pharmacology. This method was too difficult for the general pharmacologist and biological scientists because it requires the use of a complex matrix and a powerful computer to carry out the analysis. However since a more simple method was proposed by Matsumoto and Yoshimura this method has been used more frequently in the biological sciences. In this paper, the one way cumulative chi-squared statistic test and two way chi-squared statistic test are compared with the chi-squared statistic test and Wilcoxon test.
累积卡方统计量已被提出用于在各种统计模型中针对有序备择假设进行检验。与通常对有序列分类数据进行的统计检验一样，使用卡方检验、费舍尔精确检验和威尔科克森检验。药理学研究通常通过多次给药进行。从这些研究中获得的数据称为有序分类数据。广津和渋谷提出的用于在各种统计模型中针对有序备择假设进行检验的累积卡方统计量，尽管在药理学领域具有良好的适用性，但很少被使用。这种方法对于一般药理学家和生物科学家来说太难了，因为它需要使用复杂的矩阵和功能强大的计算机来进行分析。然而，自从松本和吉村提出了一种更简单的方法后，这种方法在生物科学中被更频繁地使用。在本文中，将单向累积卡方统计量检验和双向卡方统计量检验与卡方统计量检验和威尔科克森检验进行了比较。
