The Wilcoxon signed rank test for paired comparisons of clustered data
The Wilcoxon signed rank test for paired comparisons of clustered data
复制标题
DOI：
10.1111/j.1541-0420.2005.00389.x
复制
发表时间：
2006-03-01
期刊：
影响因子：
1.9
通讯作者：
Lee, MLT
中科院分区：
文献类型：
--
作者：
Rosner, B;Glynn, RJ;Lee, MLT
Wilcoxon符号秩检验是配对数据的常用非参数检验（例如，由处理前和处理后测量组成）基于油的独立分析单元。该检验不能用于聚类数据产生的配对比较（例如，如果配对比较可用于个体的两只眼睛中的每一只）。为了结合聚类，提出了符号秩检验的随机化检验公式的一般化，其中随机化的单位在聚类水平（例如，人），而各个成对的分析单元处于聚类级别内的子单元（例如，人的眼睛）。然后导出符号秩检验统计量的调整方差估计，该估计可用于平衡（每个簇的子单元数目相同）或不平衡（每个簇的子单元数目不同）数据。具有可交换的相关结构。有或没有绑定值。结果表明，所得到的检验统计量是渐近正态的集群的数量变得很大，如果集群的大小是有界的。模拟研究是基于来自符号对数正态分布的石油模拟相关排名数据进行的。这些研究表明，与基于平均聚类差异得分的普通符号秩检验或Puri和Sell的多变量符号秩检验（1971，Nonparametric Methods in Multivariate Analysis，纽约：John Wiley）相比，对于聚类数>= 20的数据集和上级功效曲线，I类错误是适当的。最后，用两个数据集来说明该方法，（i）所有眼科数据集，其涉及视网膜色素变性（13，P）患者在经历实验性外科手术之前和之后的视网膜电图（ERG）数据的比较，以及（ii）基于RP患者营养补充剂的随机化前瞻性研究的营养数据集，其中比较了研究胶囊之外的维生素E摄入量随机化以监测营养方案的依从性。
The Wilcoxon signed rank test is a frequently used nonparametric test for paired data (e.g., consisting of pre- and posttreatment measurements) based oil independent units of analysis. This test cannot; be used for paired comparisons arising from clustered data (e.g., if paired comparisons are available for each of two eyes of in individual). To incorporate clustering, a generalization of the randomization test formulation for the signed rank test is proposed, where the unit of randomization is at the cluster level (e.g., person), while the individual paired units of analysis are at the subunit within cluster level (e.g., eye within person). An adjusted variance estimate of the signed rank test statistic is then derived, which call be used for either balanced (same number of subunits per cluster) or unbalanced (different number of subunits per cluster) data. with an exchangeable correlation structure. with or without tied values. The resulting test statistic is shown to be asymptotically normal as the number of clusters becomes large, if the cluster size is bounded. Simulation studies are performed based oil simulating correlated ranked data from a signed log-normal distribution. These studies indicate appropriate type I error for data sets with >= 20 clusters and a superior power profile compared with either the ordinary signed rank test based on the average cluster difference score or the multivariate signed rank test of Puri and Sell (1971, Nonparametric Methods in Multivariate Analysis, New York: John Wiley). Finally, the methods are illustrated with two data sets, (i) all ophthalmologic data set involving a comparison of electroretinogram (ERG) data in retinitis pigmentosa (13,P) patients before and after undergoing an experimental surgical procedure, and (ii) a nutritional data set based on a randomized prospective study of nutritional supplements in RP patients where vitamin E intake outside of study capsules is compared before and after randomization to monitor compliance with nutritional protocols.