Interactive ROC curve: sensitivity, specificity and prevalence

What changes with a test threshold, and why do predictive values depend on prevalence?

An experiment from Cyril Voyant’s course, running in your browser. Full course and slides (French) · Choose another experiment.

The relationship to understand

Se=TP/(TP+FN), Sp=TN/(TN+FP), PPV=Se·p/[Se·p+(1−Sp)(1−p)]. NPV is the negative predictive value; AUC is the area under the ROC curve.

Try it and interpret

Move the threshold to move the point on the ROC curve. Then change prevalence alone: sensitivity and specificity stay unchanged in this model, while positive and negative predictive values change.

Assumptions and limitations

Equal-variance binormal illustration, not a clinical evaluation. AUC determines neither the best threshold nor an individual probability of disease. The population and the consequences of errors matter.

Go further with the course

This page focuses on one concept. The full chapter (French) contains the reasoning, further examples, slides and FAQ. Demonstration values are computed or constructed within the stated model; they are not new research findings.