Specificity

In statistics and diagnostics: the proportion of TRUE NEGATIVES correctly identified—the ability of a test to correctly identify those who DO NOT have the condition. A test with 95% specificity correctly classifies 95 out of 100 healthy people as healthy but FALSELY classifies 5 as having the condition (false positives). Specificity complements SENSITIVITY (ability to identify true positives). As a mental model applied broadly: specificity is the precision of EXCLUSION—how accurately you can rule things OUT. In decision-making: a highly specific filter rejects few good candidates (low false positive rate) but may miss some bad ones. A highly sensitive filter catches all bad ones but also rejects some good ones. The tradeoff between sensitivity and specificity is fundamental to ALL filtering, screening, and diagnostic systems.

When to use it

When diagnostic or filtering systems need calibration between false positives and false negatives; when the cost asymmetry between error types needs to drive system design; when understanding why tests that are sensitive aren't necessarily specific; when screening systems need explicit tradeoff decisions.

How it can help

When designing any filtering or diagnostic system: explicitly choose the sensitivity-specificity tradeoff based on the COSTS of each error type. High specificity (prioritize when false positives are costly): hiring (false positives = bad hires are expensive), medical treatment (false positives = unnecessary surgery), spam filtering (false positives = losing important email). High sensitivity (prioritize when false negatives are costly): disease screening (false negatives = missed cancer), safety systems (false negatives = undetected threats). The practice: for every filter you use, ask: 'which error is MORE costly—false positive or false negative?'—then calibrate accordingly.