Accumetric Test Performance Analysis software, receiver operating characteristic analysis, ROC analysis

Accumetric screen of the ROC analysis of a single curve. Note the name of the test, the name of the disorder, the sample size, the number and prevalence of abnormal and normal cases, the number of unique cutpoints found in the data, the area under the ROC curve, its asymptotic standard error, its asymptotic 95% confidence interval, and the statistical significance of the difference from the line of no information. The line of no information is the ROC curve with area 0.5 of a test operating no better than random. Note how asymptotic standard errors can sometimes yield confidence intervals whose upper bound exceeds 1. This is why Accumetric also provides bootstrap confidence intervals, which are range-respecting.