Base-Rate Neglect
The tendency to overweight specific, story-shaped evidence about an individual case while underweighting the underlying frequency in the population.
Maya Bar-Hillel's 1980 Acta Psychologica paper established base-rate neglect as a robust pattern across dozens of probability judgment experiments. The classic illustration is the cab problem: 85% of cabs in a city are green and 15% are blue; a witness identifies a hit-and-run cab as blue and is right 80% of the time. The probability the cab was actually blue is around 41% — the witness's 80% accuracy collides with the 15% prior. Most respondents answer 80%, ignoring the base rate entirely. The same pattern shows up in medical screening (positive test results on rare conditions are mostly false positives), hiring decisions (memorable interview moments outweigh selection rates), and security profiling (vivid suspect features overwhelm the underlying frequency of guilt). Bayesian reasoning is the formal corrective: the posterior probability has to be weighted by the prior, not just the diagnostic strength of the evidence.
Frequently Asked Questions
What is a real-world cost of base-rate neglect?
Medical screening of rare diseases is the canonical example. A test that is 99% accurate for a condition that affects 1 in 10,000 people will generate roughly 100 false positives for every true positive. Patients and clinicians who anchor on "99% accurate" without checking the base rate routinely overestimate the meaning of a positive result.
How do you guard against base-rate neglect in decisions?
Pull the base rate first, before looking at the case-specific evidence. Write the prior probability down. Then update from it, rather than starting from the vivid story and adjusting toward the base rate. The order matters: the brain anchors on whichever number it sees first, and stories are stickier than statistics.
Is base-rate neglect the same as the conjunction fallacy?
Different but related. The conjunction fallacy (the famous "Linda is a feminist bank teller" problem) is judging a compound event as more probable than one of its parts. Base-rate neglect is a broader failure: ignoring how rare the underlying category is. Both stem from substituting representativeness for probability.