Title: A comparative study of methods for testing the equality of two or more ROC curves
Authors: Aris Fanjul Hevia - Universidad de Santiago de Compostela (Spain) [presenting]
Wenceslao Gonzalez-Manteiga - University of Santiago de Compostela (Spain)
Abstract: The problem of comparing the accuracy of diagnostic tests is usually carried out through the comparison of the corresponding ROC curves. This matter has been approached from different perspectives. Usually, ROC curves are compared from their respective AUCs, but in the case where there is no uniform dominance between the involved curves other procedures are preferred, such as the ones based on the distance of two empirical quantil processes, or the ones that take advantage of the fact that the ROC curves can be viewed as cumulative distribution functions. Although the asymptotic distributions of the statistics behind these methods are, in general, known, resampling plans are considered. With the purpose of comparing the performance of these different approaches, a simulation study is carried out. Additionally, some of these methods are extended to the case where covariates are available.