Skewed distributions and inferences concerning quantiles are common in the health and social science researches. And most standard simultaneous inference procedures require the normality assumption. For example, few methods exist for comparing the medians of independent samples or quantiles of several distributions in general. To our knowledge, there is no easy-to-use method for constructing simultaneous confidence intervals for multiple contrasts of quantiles in a one-way layout. In this paper, we develop an asymptotic method for constructing such intervals both for differences and ratios of quantiles and extend the idea to that of right-censored time-to-event data in survival analysis. Small-sample performance of the proposed method and a bootstrap method were assessed in terms of coverage probabilities and average widths of the simultaneous confidence intervals. Good coverage probabilities were observed for most of the distributions considered in our simulations. The proposed methods have been implemented in an R package and are used to analyze two motivating datasets.

中文翻译：

---

分位数对比的同时置信区间

关于分位数的偏态分布和推断在健康和社会科学研究中很常见。并且大多数标准的同时推理程序都需要正态性假设。例如，通常很少有方法用于比较独立样本的中位数或几个分布的分位数。据我们所知，没有一种易于使用的方法可以为单向布局中的多个分位数对比同时构建置信区间。在本文中，我们开发了一种渐近方法，用于为分位数的差异和比率构建此类间隔，并将该想法扩展到生存分析中右删失的事件时间数据的想法。在覆盖概率和同时置信区间的平均宽度方面评估了所提出方法和引导方法的小样本性能。在我们的模拟中考虑的大多数分布都观察到了良好的覆盖概率。所提出的方法已在 R 包中实现，并用于分析两个激励数据集。