A0870
Title: Nonparametric comparisons of multiple distributions under uniform stochastic ordering with survival data
Authors: Chuan-Fa Tang - University of Texas at Dallas (United States) [presenting]
Abstract: When survival data are collected from ordered hazardous environments, uniform stochastic ordering, also known as hazard rate ordering for absolutely continuous distributions, is applied to describe the relation of distributions and achieve better efficiency in statistical inferences. Equality tests under HRO and goodness-of-fit tests are discussed for the adequacy of HRO in the present data with a nonparametric ordinary dominance curve (ODC). A grouping method is proposed based on hazard, and inferences on the ODC and receiver operating characteristic curve (ROC) are developed. Simulation results are provided to support the theoretical findings, and the proposed methods are illustrated for the data on premature infants.
