Tarone R E
Biometrics. 1982 Jun;38(2):457-62.
A method is developed for the use of historical control information in testing for a trend with Poisson data. The resulting test statistic is a modification of a statistic proposed by Armitage (1955, Biometrics 11, 375-386) to test for a trend in frequencies. The contribution of the historical control data in the analysis depends upon the degree of heterogeneity of the historical counts.
开发了一种在泊松数据趋势检验中使用历史对照信息的方法。所得的检验统计量是对阿米蒂奇(1955年,《生物统计学》11卷,375 - 386页)提出的用于检验频率趋势的统计量的一种修正。历史对照数据在分析中的贡献取决于历史计数的异质性程度。