Confirmation Bias: The Brain's Stubborn Preference for Being Right

Wason's Card Task

In 1960, psychologist Peter Wason devised one of the most famous experiments in the history of cognitive psychology. He presented participants with four cards showing E, K, 4, and 7, along with the rule: "If a card has a vowel on one side, it has an even number on the other." Participants were asked which cards to turn over to test whether the rule was true.

The correct answer is E and 7 — you need to confirm vowels have even numbers, and verify odd numbers don't have vowels. Most participants chose E and 4. They selected the confirming cards rather than the potentially disconfirming one. Turning over 4 cannot disprove the rule regardless of what is on the other side. Only 7 could reveal a violation. People systematically avoided the card that could prove them wrong.

Motivated Reasoning

Confirmation bias is not simply a failure of logic. It is a motivated process. When people hold a belief, they experience discomfort from information that challenges it — the same cognitive dissonance Festinger described — and satisfaction from information that supports it. The result is that the search for evidence is never neutral. People ask different questions, apply different levels of scrutiny, and remember different things depending on whether incoming information supports or threatens their existing view.

In a series of studies by Lord, Ross, and Lepper in 1979, participants who held strong views about capital deterrence were given two studies — one supporting deterrence, one opposing it. Both studies were methodologically equivalent, but participants rated the study supporting their prior belief as significantly more rigorous and persuasive. After seeing both studies, they ended the experiment more confident in their original positions than before. Exposure to balanced evidence made beliefs more extreme, not less.

Why Expertise Doesn't Protect You

Confirmation bias is not reduced by intelligence or expertise — in some cases it is amplified. Intelligent people are simply better at rationalizing. They can generate more sophisticated arguments for their existing beliefs and construct more persuasive accounts for why contrary evidence is flawed. Studies of medical diagnosis show that physicians who form an early hypothesis tend to gather confirming evidence and anchor on the initial impression, even when new symptoms do not fit.

The structural solution, as with the halo effect, is procedural. Scientific peer review, adversarial collaboration between researchers with opposing hypotheses, pre-registration of study designs, and formal devil's advocate practices in decision-making all exist specifically to make confirmation bias harder to act on. The bias does not disappear, but the environment can be designed to reduce its consequences.