一个用于衡量相对于基线变化的汇总统计量。

A summary statistic for measuring change from baseline.

作者信息

Donahue R M

机构信息

GlaxoWellcome, Research Triangle Park, North Carolina 27709, USA.

出版信息

J Biopharm Stat. 1997 May;7(2):287-99. doi: 10.1080/10543409708835187.

DOI:10.1080/10543409708835187

PMID:9136070

Abstract

A statistic, W, for measuring change from baseline is developed. Its distribution is found. Simulations using W and analysis of covariance (ANCOVA) are run and the results are compared. W is found to be less powerful than ANCOVA, yet is not seen to suffer some of the ill effects to which ANCOVA can fall prey, namely bias introduced by chance covariate imbalance.

摘要

开发了一种用于衡量相对于基线变化的统计量W。确定了其分布。使用W进行模拟并进行协方差分析（ANCOVA），然后比较结果。发现W的功效不如ANCOVA，但未发现它会遭受ANCOVA可能遭遇的一些不良影响，即偶然协变量不平衡所引入的偏差。