Evidence-appraisal glossary
Discrimination
A test or model's ability to assign higher risk scores to people who have the condition than to those who do not; it is usually summarized by the C-statistic or area under the ROC curve.
Also called: discriminative ability, discriminatory power.
Discrimination answers a ranking question: if you pick one person with the disease and one without, how often does the model rate the diseased person as higher risk? It is reported as a value from 0.5 (no better than chance) to 1.0 (perfect separation). Discrimination is not the same as calibration; a model can rank people well yet still output probabilities that are systematically too high or too low, so both properties should be checked.
This is a plain-language methodology definition for reading research. It is general education, not medical advice.