Specificity
Definition
Specificity measures how many truly negative cases are correctly recognized as negative.
Specificity is the true negative rate TN/(TN+FP): the proportion of truly negative cases correctly recognized as negative at a given threshold.
Simply explained
Of all cases without the target feature, it counts the proportion where the system correctly gives an all-clear.
Why it matters in practice
The term helps to select, develop, evaluate AI systems in a well-founded manner and realistically assess their technical risks.
Example from business
A fraud detection system measures how many legitimate transactions are correctly not flagged.
Limits and pitfalls
- Reporting specificity alone without sensitivity, threshold, and target population, or confusing it with negative predictive value and class prevalence.
Origin & history
The metric originates from diagnostic test theory and is used together with sensitivity to evaluate binary decisions.
Use in mid-sized companies
- Possible application: A fraud detection system measures how many legitimate transactions are correctly not flagged.
Frequently asked questions
What does specificity mean?
Specificity is the true negative rate TN/(TN+FP): the proportion of truly negative cases correctly identified as negative at a given threshold. What mistakes should be avoided? Reporting specificity alone without mentioning sensitivity, threshold, and target population, or confusing it with negative predictive value and class prevalence.
Sources
- Sensitivity and Specificity · NIST SEMATECH e-Handbook, Sensitivity and Specificity
Source checked on 10.10.2026. Editorial application examples are hypothetical; no evidence of real customer results.
Editorially reviewed on: