BiasesBelief & evidence
Confirmation Bias
Also called Myside bias · Biased assimilation
Seeking, noticing, and remembering evidence that supports what you already believe.
Definition
Confirmation bias is the habit of treating evidence for your view more kindly than evidence against it.
It shows up at every stage: in which questions you ask, which sources you open, what you notice, how hard you check, and what you remember later. Psychologist Raymond Nickerson called it “ubiquitous” in his 1998 review of the research. Intelligence doesn’t protect you from it. Skilled reasoners are often better at explaining away inconvenient evidence.
How it works
The bias rarely feels like bias. It feels like being reasonable. Evidence that fits your view seems obviously sound, so you accept it quickly. Evidence that doesn’t fit seems suspicious, so you look for flaws, and you usually find some, because most evidence has some. The result is a double standard you don’t notice applying.
In Peter Wason’s 1960 “2-4-6” task, people were told that the sequence 2-4-6 followed a rule and were asked to find the rule by testing sequences of their own. Most tested only sequences that fit their guess, like 8-10-12, and rarely tried ones that might prove it wrong. The real rule was simply “any increasing numbers,” and many never found it.
How to spot it
You remember the hits and forget the misses (“I always get a call right after I think of someone”).
You search for “evidence that X,” not “is X true?”
A study you like gets a quick read, and a study you dislike gets a careful one.
You can explain away every counterexample, but none of your supporting examples.
You can’t say what evidence would change your mind.
Examples
01Two groups, the same studies
In a 1979 Stanford experiment, supporters and opponents of the death penalty read the same two studies, one suggesting it deters crime and one suggesting it doesn’t. Each group rated the study that agreed with them as more convincing, and both left more certain of their original views.
What went wrong
The same mixed evidence pushed two groups further apart. Each judged the methods by whether they liked the conclusion.
02The new-car test drive
After buying a car, an owner starts reading glowing reviews of the model and skips the negative ones as written by people “who don’t get it.”
What went wrong
The decision is already made, so evidence is now being collected to protect it instead of to judge it. Reading only praise feels like research but doesn’t test anything.
03The search bar
Someone worried about a health symptom searches “can [symptom] mean [serious disease]” and finds pages that say it can.
What went wrong
The query was built to confirm. A search like “most common causes of [symptom]” would have turned up the much likelier, ordinary explanations.
Why we do it
Being wrong is uncomfortable, especially about beliefs tied to identity or to people we trust. Checking every belief fairly would also be exhausting, so the mind defaults to keeping what it has. Some researchers argue the habit makes sense for groups: in a debate, each side’s bias pushes it to find the best case for its view, and the group as a whole may reason better than any one person. That only works when the other side is actually in the room.
Check yourself
1Name what would change your mind
Before looking at evidence, decide what result would count against your view. If nothing would, you’re not testing it.
What would I expect to see if I were wrong, and have I actually looked for it?
2Apply the swap test
Imagine the same study, source, or anecdote pointed the other way. Would you judge it the same?
If this came out the other way, would I trust the method just as much?
3Search for the strongest disagreement
Go looking for the best argument against your view, not the weakest.
Who’s the smartest person who disagrees with me, and what’s their best argument?
When the shortcut is reasonable
Not every belief needs a fresh audit. If you have strong, well-tested reasons for a view, it’s rational to want more evidence before giving it up because of one odd result. A single surprising study shouldn’t erase a large body of research. The bias is in the double standard: demanding more of the evidence you dislike than of the evidence you like.
Sources & further reading
- Raymond S. Nickerson, “Confirmation Bias: A Ubiquitous Phenomenon in Many Guises,” Review of General Psychology 2, no. 2 (1998): 175–220.
- P. C. Wason, “On the Failure to Eliminate Hypotheses in a Conceptual Task,” Quarterly Journal of Experimental Psychology 12 (1960): 129–140.
- Charles Lord, Lee Ross and Mark Lepper, “Biased Assimilation and Attitude Polarization,” Journal of Personality and Social Psychology 37, no. 11 (1979): 2098–2109.
- Hugo Mercier and Dan Sperber, The Enigma of Reason (Harvard University Press, 2017).
End of entry · Back to top