Cobra Effect · Thinking and evidence
Confirmation bias
We look for evidence that agrees, and rarely test what could say no.
6 cards, read aloud in 2:31, with a test and sources.
Three numbers on a card. 2, 4, 6. They follow a rule. Find it.
London, 1960. The psychologist Peter Wason sits students down with that card. You may write down any three numbers you like, and he will tell you whether they fit the rule. Try as many as you want. When you are sure, announce the rule.
Almost everyone does the same thing.
They think, going up in twos. So they try 8, 10, 12. Yes, says Wason. Then 20, 22, 24. Yes. Then 100, 102, 104. Yes. They announce the rule with confidence. Going up in twos. Wrong.
The rule was any three numbers that go up.
1, 2, 3 would have fitted. So would 5, 40, 900. But almost nobody tried those. Every test they ran was designed to hear yes. Not one was designed to hear no, and no was the only answer that could have taught them anything.
That is confirmation bias.
We look for evidence that we are right and treat it as proof. We do not go looking for the evidence that would show we are wrong. Francis Bacon described it in 1620. Once the mind adopts an opinion, it draws everything else in to support it. Wason gave it the name, and the numbers.
It does not feel like bias from the inside. It feels like research.
The doctor who orders the test that would confirm her first guess. The investor who reads every article that agrees with him and calls it due diligence. The detective who has a suspect and finds that everything fits. Each is collecting yes after yes, like the students with their fence posts.
The fix is Wason’s own. Try to hear no.
Before you announce your rule, write down what you believe. Then design one test that would fail if you are right. If you cannot think of one, you do not have a belief. You have a habit. Say 1, 2, 3 out loud, and see what the world says back.
Sources
- Peter Wason, Wikipedia. The man and his three famous tasks. The 1960 paper is called On the failure to eliminate hypotheses in a conceptual task, in the Quarterly Journal of Experimental Psychology. Search the title.
- Mistakes Were Made, but Not by Me, Carol Tavris and Elliot Aronson, 2007. How people defend a belief once they hold it, from doctors and detectives to whole governments. Readable, alarming and full of documented cases.
- Confirmation bias, Wikipedia. The experiments since Wason, the different forms the bias takes, and the argument about whether it is a flaw or a feature of how we reason together.
Nearby ideas
- Survivorship bias. The failures you never see can reverse the lesson you draw.
- Falsifiability. A claim that nothing could disprove tells you nothing.
- The narrative fallacy. We invent causes for events because stories feel like explanations.
- The planning fallacy. Why projects nearly always take longer than we plan.
- Base rate neglect. A good test for a rare thing still gives mostly false alarms.
- Anchoring. The first number you hear bends every estimate that follows.
- The availability heuristic. Vivid stories make rare dangers feel common.
- Hindsight bias. Once you know the ending, it always looks obvious.