Confirmation bias is the tendency to favor evidence that fits what you already believe. It shapes what you look for, how you read what you find and what you remember afterward, and the evidence against your view gets less attention and less weight. It does not depend on wanting the belief to be true. A guess points you at the cases it predicts, and checking those is easier than imagining the rest.

Three places it happens

Our guide to confirmation bias splits it into three errors at different points:

  • The search. You look for cases that would fit the belief and skip the ones that would break it.

  • The reading. Evidence that fits is taken at face value. Evidence that does not fit is checked for flaws or explained away.

  • The memory. Fitting cases are easier to recall later, so the belief seems better supported than it ever was.

A 1998 review by Raymond Nickerson treated all three as one phenomenon and described it as ubiquitous.

The experiment behind the idea

In 1960 Peter Wason published a study in which 29 undergraduates were told that the experimenter had a rule for sets of three numbers and that 2, 4, 6 fit it. They proposed triples, heard yes or no for each, and announced the rule once confident.

The rule was any three numbers in ascending order. Only 6 of the 29 announced it without first announcing a wrong one. Participants tended to test triples that their own guess predicted, collect a run of yeses and then state the guess with confidence. Hardly anyone tried a triple meant to draw a no.

The page has a version of the task that takes about 2 minutes and scores how you tested. For the research in more depth, the free mini course on confirmation bias takes 12 to 15 minutes.

Everyday examples

  • You suspect a food upsets your stomach and start a diary. The rough days that follow the food get written down. Rough days without it, and good days with it, never make the page.

  • A manager decides early that a new hire is a weak performer. A missed deadline then counts as proof, and a success gets put down to luck or a teammate.

  • You search online for whether a supplement works. A query phrased as "does X help with Y" brings back pages that say yes, and those get read first.

  • Two friends read the same news story about a politician they disagree on. Each finds their existing view confirmed.

Related patterns, such as survivorship bias and the halo effect, have their own entries in the cognitive biases library.

How to catch it

The methods that work change the procedure. In a 1984 study by Lord, Lepper and Preston, asking people to imagine how they would have judged a study if it had reached the opposite result reduced biased reading. Asking them to be fair did not.

  1. 1.

    Name one observation your belief says you will never see, then go and look for it.

  2. 2.

    Consider the opposite: judge the evidence as if it had pointed the other way.

  3. 3.

    Write your prediction down before the evidence arrives.

  4. 4.

    Count the misses, including the times the cause was present and the effect was absent.

  5. 5.

    Give one person, or one hour, the job of building the strongest case against your preferred option.

For practice on a belief of your own, the guide points to a Critical Thinking Test and a Check the Facts worksheet.

From the Psychology.com editorial team.