Cognitive Biases
Your brain takes shortcuts. Most of the time they work. The rest of the time they steer you wrong in predictable, named ways — and you can learn to catch yourself doing it.
What a bias actually is
A cognitive bias is a systematic — not random — error in how the mind processes information. The brain runs on fast, low-effort rules of thumb called heuristics (Kahneman & Tversky, 1974). Heuristics are usually good enough; biases are the price you pay when they aren't.
You can't delete a bias. You can recognize when conditions favor one, slow down, and reach for a corrective procedure (a checklist, a calculation, a second opinion).
Two systems, one mind
Kahneman's System 1 is fast, automatic, and effortless — it produces hunches. System 2 is slow, deliberate, and effortful — it does math. Biases live in System 1; the cure is usually to wake System 2 up.
That's why almost every fix on this page looks like a small extra step: write the number down, compare against a base rate, list the opposite, ask a colleague before you anchor.
Filter the catalog
Anchoring lab — feel the pull
The classic Tversky & Kahneman experiment
In 1974, a roulette wheel was rigged to stop on either 10 or 65. Subjects watched it spin, then estimated the percentage of African countries in the UN. Those who saw 10 guessed a median of 25%. Those who saw 65 guessed a median of 45%. The wheel's number — obviously random — pulled their estimate toward it.
Type your guess, then lock it in to see how far it sits from the anchor and from the real answer.
Base-rate trainer
Why "1% disease + 95% accurate test" is not what you think
You test positive for a disease that affects 1% of the population. The test correctly flags sick people 95% of the time (sensitivity) and correctly clears healthy people 95% of the time (specificity). What's the chance you actually have the disease?
The lesson: when the base rate is low, even a very accurate test produces mostly false alarms. The intuitive 95% is wrong — most positive tests are well people.
Spot the bias
Read each scenario. Pick the bias most clearly at work. The catalog above defines every option.
Calibration trainer — fight overconfidence
How well do you actually know what you think you know?
For each question, give a 90% confidence interval — a range you're 90% sure contains the true answer. The wider the range, the more likely you'll be right; but a useless-wide range tells you nothing. If you're well-calibrated, the true answer lands inside your range about 9 times out of 10. Most people are over-confident (hit rate under 70%) and don't notice until they track it.
How to use it: do at least 10 questions to get a meaningful sample. If your hit rate sits below 90%, you are systematically overconfident — widen your ranges. If it sits at 100%, your ranges are uselessly wide — narrow them.
Connect the dots
Quiz
15 questions on cognitive biases.
Flashcards
Tap to flip. Twenty-four essential biases.
Teacher mode
Lesson outline, quick-reference card, and a printable worksheet with answer key.
Lesson outline (45 min)
- 5 min · Hook — Run the anchoring demo live: ask half the class for the last 2 digits of their phone number, then have everyone estimate the number of countries in Africa. Plot. The "high digits" half almost always estimates higher.
- 10 min · Concept — System 1 / System 2. Why heuristics exist (energy economy). The difference between random error (noise) and systematic error (bias).
- 10 min · Catalog — Walk the six families: memory, social, probability, belief, ego-protective, decision. Give one fresh example for each.
- 15 min · Practice — Spot-the-bias scenarios from this module's catalog. Have students argue for two competing labels (e.g. confirmation vs. anchoring) and defend.
- 5 min · Wrap — Each student writes one bias they recognize in their own thinking this week, plus the specific countermeasure they'll try.