What is a Randomization Test?
What is a Randomization Test?
复制标题
DOI：
10.1080/01621459.2023.2199814
复制
发表时间：
2022-03
影响因子：
3.7
通讯作者：
Yao Zhang;Qingyuan Zhao
中科院分区：
文献类型：
--
作者：
Yao Zhang;Qingyuan Zhao
摘要 上个世纪，随机测试的含义在统计教育和实践中变得模糊。本文对统计学的这一核心概念进行了新的尝试。引入了一个新术语“准随机化检验”来定义基于理论模型的显着性检验，并将这些检验与基于随机化物理行为的“随机化检验”区分开来。通过真正的阶梯楔形整群随机试验说明了这种区别的实际重要性。基于最近关于随机化推理的文献，开发了条件随机化测试的一般框架，并给出了一些构造条件事件的实用方法。然后应用所提出的术语和框架来理解几种广泛使用的（准）随机化检验，包括费舍尔精确检验、治疗效果的排列检验、独立性和条件独立性的准随机化检验、自适应随机化和保形预测。本文的补充材料可在线获取。
Abstract The meaning of randomization tests has become obscure in statistics education and practice over the last century. This article makes a fresh attempt at rectifying this core concept of statistics. A new term—“quasi-randomization test”—is introduced to define significance tests based on theoretical models and distinguish these tests from the “randomization tests” based on the physical act of randomization. The practical importance of this distinction is illustrated through a real stepped-wedge cluster-randomized trial. Building on the recent literature on randomization inference, a general framework of conditional randomization tests is developed and some practical methods to construct conditioning events are given. The proposed terminology and framework are then applied to understand several widely used (quasi-)randomization tests, including Fisher’s exact test, permutation tests for treatment effect, quasi-randomization tests for independence and conditional independence, adaptive randomization, and conformal prediction. Supplementary materials for this article are available online.