Test Sensitivity vs. Specificity

Sensitivity is the probability that a test correctly identifies someone WITH a condition (true positive rate). Specificity is the probability that a test correctly identifies someone WITHOUT the condition (true negative rate). A highly sensitive test rarely misses cases (few false negatives) but may flag healthy people (false positives). A highly specific test rarely flags healthy people (few false positives) but may miss some cases (false negatives). Practitioners learn to match test choice to clinical context: use sensitive tests for screening (you don't want to miss anyone) and specific tests for confirmation (you want to be sure before treating). The deep practitioner insight is that sensitivity and specificity are properties of the TEST, but what patients and doctors actually need to know is predictive value — which depends on the BASE RATE (prevalence) of the condition.

When to use it

When evaluating or designing any detection, screening, or classification system; when interpreting test results (medical, security, quality); when deciding whether to use a broad filter or narrow filter; or when communicating the meaning of positive or negative test results to non-experts.

How it can help

This framework applies to any detection or classification system. In hiring, a highly 'sensitive' screen (accepting anyone who might be good) lets in many false positives. A highly 'specific' screen (only accepting candidates who are definitely good) misses many qualified people. In fraud detection, cybersecurity, and quality control, the sensitivity-specificity tradeoff determines whether your system catches threats (at the cost of false alarms) or minimizes false alarms (at the cost of missed threats). Understanding that a test's real-world usefulness depends on the base rate of what you're looking for — not just the test's accuracy — prevents enormous waste.