Best D J, Rayner J C
CSIRO, IAPP Biometrics Unit, North Ryde, Australia.
Biometrics. 1991 Jun;47(2):589-93.
Tango (1984, Biometrics 40, 15-26) proposed a clustering index for testing for clusters of disease in time. A test based on this clustering index was shown to compare favourably with other statistical tests. In this article we show that Pearson's X2 and its components perform well in testing for clusters of disease in time. The rth of these components identifies a departure from uniformity in moments up to the rth and so helps describe the alternative, if any, to uniformity.
坦戈(1984年,《生物统计学》第40卷,第15 - 26页)提出了一种用于检测疾病随时间聚类的聚类指数。基于该聚类指数的检验与其他统计检验相比表现良好。在本文中,我们表明皮尔逊卡方及其分量在检测疾病随时间聚类方面表现良好。这些分量中的第r个分量可识别直至第r阶矩偏离均匀性的情况,从而有助于描述与均匀性相对的情况(如果存在的话)。