Cognitive bias

A cognitive bias is a systematic, predictable pattern of deviation from rational judgment in human reasoning under uncertainty. Identified empirically in decision-science research since the 1970s, biases arise from the heuristic shortcuts the mind uses to make rapid judgments and produce reliable errors in identifiable circumstances.

Cognitive bias matters because the same shortcuts that let us make rapid, mostly-good decisions in everyday life produce predictable, systematic errors in identifiable circumstances. The errors are not random; they are reproducible across populations, decades, and cultures. Knowing the patterns is the first prerequisite for noticing them in your own thinking. The biases also matter because they aggregate across institutions: the cognitive errors of doctors, judges, hiring managers, and policymakers compound into population-level outcomes that no individual would consciously endorse.

The technical concept of cognitive bias was articulated in Amos Tversky and Daniel Kahneman's foundational 1974 Science paper, "Judgment under uncertainty: Heuristics and biases." Their core argument was that people make probabilistic judgments not by computing posterior probabilities but by relying on a small set of mental shortcuts (heuristics) that work well most of the time but produce systematic errors in identifiable circumstances. Kahneman's 2002 Nobel Prize anchored this research as one of the most consequential in modern psychology. The field has since grown to encompass dozens of distinct biases, with Baron's (2008) compendium listing 53 and contemporary lists exceeding 200 (most of which are minor variants of a smaller set of fundamental patterns).

Susceptibility to cognitive biases predicts real-world outcomes — financial decision quality, medical decision quality, voting behavior, susceptibility to misinformation. People who score better on rationality tests make measurably better decisions in domains as varied as retirement saving and medical adherence (Bruine de Bruin, Parker, & Fischhoff, 2007). Crucially, susceptibility to bias is largely independent of cognitive ability. Stanovich's research program over three decades has shown that people with high IQs are not meaningfully less susceptible to most cognitive biases — a finding he labels dysrationalia: the inability to think rationally despite having adequate intelligence.

Cognitive bias is not the same as social bias or prejudice. The technical term refers specifically to systematic deviations from rational judgment under uncertainty — not to discrimination based on protected characteristics. Cognitive biases also are not always failures: many of them are adaptive shortcuts that produce good-enough answers under time pressure. The biases become problematic when the stakes are high, time is available, and the heuristic answer differs from the deliberative one. Awareness of bias does not eliminate it; debiasing requires structured practice, not just insight.

The Cognitive Bias Susceptibility tool measures eight specific biases (anchoring, availability, base-rate neglect, framing, hindsight, overconfidence, sunk-cost, confirmation bias) using validated task structures from the Heuristics-and-Biases Inventory (Berthet, 2023), the Adult Decision-Making Competence battery (Bruine de Bruin et al., 2007), and the Comprehensive Assessment of Rational Thinking (Stanovich, West, & Toplak, 2016). See the tool methodology page.

• Heuristic
• Anchoring effect
• Confirmation bias
• Bias blind spot