Confirmation Bias

What is Confirmation Bias?

Confirmation bias is the tendency to search for, interpret, favor, and recall information in ways that confirm pre-existing beliefs or hypotheses. It can affect the questions people ask and the evidence they seek or interpret. A hiring manager who believes a candidate is strong may unconsciously focus on interview answers that support that impression while discounting signs of weakness. Confirmation bias can overlap with motivated reasoning, but it does not require a conscious wish to defend a preferred answer.

How it works

Peter Wason identified confirmation bias experimentally in 1960 using what became known as the 2-4-6 task. Participants were told that the number sequence 2, 4, 6 conformed to a rule, and they had to discover it by proposing new sequences. Testing only triples that fit a narrow hypothesis can leave broader rules unexamined. If you suspect “ascending even numbers,” another even-number sequence may confirm your guess without distinguishing it from a wider rule. The target rule accepted any three numbers in increasing order. Only six of the 29 participants identified it without first announcing an incorrect rule; this does not mean that the others never discovered it. Raymond Nickerson's comprehensive 1998 review later documented how this same pattern manifests across scientific reasoning, legal judgment, medical diagnosis, and everyday belief formation.

Confirmation Bias Example: The Mayfield Fingerprint Case

The Brandon Mayfield fingerprint case illustrates the risk in expert evidence assessment. In 2004, Mayfield was wrongly identified as the source of a fingerprint connected to the Madrid train bombings, detained as a material witness, and released after Spanish investigators identified another person. The U.S. Department of Justice Inspector General found that genuine similarities helped start the error, but examiners then interpreted ambiguous details in the questioned print using Mayfield’s known print. They also failed to give adequate weight to the Spanish investigators’ disagreement. The report identified several contributing factors. It is a documented case of a failed examination process, not a controlled experiment isolating one bias.

Why it matters

Confirmation bias threatens the quality of decisions wherever evidence must be evaluated — in hiring, investment analysis, product research, medical diagnosis, and policy design. Explicit alternative hypotheses and a record of contradictory evidence are practical safeguards to consider. Their effectiveness depends on the task and how they are used; awareness alone is not a proven solution.

Myside bias focuses on evaluating evidence in ways that favor one's own existing views.

Sources and further reading

• Wason, P. C. (1960): On the Failure to Eliminate Hypotheses in a Conceptual Task
• Nickerson, R. S. (1998): Confirmation Bias: A Ubiquitous Phenomenon in Many Guises

Case source: Report to Congress on Implementation of Section 1001 of the USA PATRIOT Act: Brandon Mayfield matter.