Abstract

A novel covering principle was recently proposed to address the multiplicity issue in the multiple hypotheses testing with a hierarchical structure. In this article, a sieve method is proposed to decompose the family of hypotheses into subfamilies. A Shiny web app in R is developed which makes it easier for users to build the objective-tailored multiple testing procedure based on the covering principle by a smart phone or a tablet with a web browser. The proposed methods are applied to two real clinical trials, demonstrating the flexibility and greater power than the popular gate-keeping method.

中文翻译：

---

覆盖原理在多目标临床试验中的应用

摘要

最近提出了一种新颖的覆盖原则来解决具有层次结构的多重假设检验中的多重性问题。在本文中，提出了一种筛法将假设族分解为子族。开发了 R 中的 Shiny Web 应用程序，使用户可以更轻松地通过带有 Web 浏览器的智能手机或平板电脑构建基于覆盖原理的目标定制的多重测试程序。所提出的方法应用于两个真实的临床试验，证明了比流行的把关方法更灵活和更强大的能力。