A simulation based study for comparing tests associated with Receiver Operating Characteristic (ROC) curve

Abstract

Receiver Operating Characteristic curves and the Area Under Curve (AUC) are widely 7 used to evaluate the predictive accuracy of diagnostic tests. The parametric methods of 8 estimating AUCs are well established while nonparametric methods, such asWilcoxon’s 9 method, lack proper research. This study considered three standard error techniques, 10 namely, Hanley and McNeil, Hanley and Tilaki, and DeLong methods. Several param11 eters were considered, while measuring the predictor on a binary scale. The normality 12 and type I error rate was violated for Hanley andMcNeil’s method while asymptotically 13 DeLong’smethod performed better.Hanley and Tilaki’s Jackknifemethod andDeLong’s 14 method performed equally well.