时 间：2023年05月05日10:00-11:00
地 点：理科大楼A1714
报告人：Jesse Frey维拉诺瓦大学教授
主持人：徐进 华东师范大学教授
摘 要：
Melded confidence intervals were proposed by Fay, Proschan, and Brittain (2015, Biometrics) as a way to combine together two independent one-sample confidence intervals to obtain a two-sample confidence interval for a quantity like a difference or a ratio. Simulation-based and computation-based work has suggested that melded confidence intervals always provide at least the nominal coverage. However, we show here that for the case of melded confidence intervals for a difference in population quantiles, the confidence intervals do not guarantee the nominal coverage. We derive a lower bound on the coverage for a one-sided melded confidence interval, and we show that there are pairs of distributions that make the coverage arbitrarily close to this lower bound. One specific example of our numerical results is that the 95% melded upper bound on the difference between two population medians offers a guaranteed coverage of only 88.3% when both samples are of size 20. This work is joint work with my Villanova University colleague Yimin Zhang (张一旻)。
报告人简介：
Jesse Frey is a professor in the Department of Mathematics and Statistics at Villanova University in Villanova, Pennsylvania, USA. He earned his PhD in statistics from The Ohio State University in 2005. His research areas include nonparametric statistics and ranked-set sampling.