Online Program Home
My Program

Abstract Details

Activity Number: 201
Type: Contributed
Date/Time: Monday, August 1, 2016 : 10:30 AM to 11:15 AM
Sponsor: Section on Nonparametric Statistics
Abstract #321578
Title: AUC Regression for Multiple Comparisons of Monotone Zero-Dose Control Experiments
Author(s): Johanna Van Zyl* and Jack D. Tubbs
Companies: Baylor University and Baylor University
Keywords: Zero-dose Control ; Multiple comparisons ; Non-parametric ; Mann-Whitney ; ROC
Abstract:

Dodd and Pepe (2003) proposed a semi-parametric regression model for the area under the ROC curve (AUC). Their method permitted covariate adjustments for the non-parametric Mann-Whitney test statistic. Buros et al. (2015) used a method suggested by Zhang et al. (2011) to test for treatment differences under the ordered alternative hypotheses as considered by the Jonckheere trend test. Buros et al. (2015) further developed a multiple comparison procedure in which she found all breaks in the ordered alternative hypothesis of the Jonckheere trend test. In this paper the method given in Buros et al. (2015) is modified so as to provide a non-parametric alternative to Shirley (1977) where each treatment arm is compared to the zero-dose control. The proposed method is evaluated in a Monte Carlo simulation study demonstrating that the familywise error rate is controlled and the proposed method is competitive with Shirley.


Authors who are presenting talks have a * after their name.

Back to the full JSM 2016 program

 
 
Copyright © American Statistical Association