Pick an opinion you actually hold, something you feel reasonably sure about. Could be political, could be about diet, could be about whether remote work makes people lazy. Now, honestly: when was the last time you went looking for the strongest argument against it? Not a strawman version you could knock down, but the best case the other side has to offer.
For most of us, the honest answer is something like “a while ago” or “never.”
A quick note before we go on. I’m a curious generalist here, not a psychologist, and what follows is reading and reflection on the research, not advice about your own mind. The studies I mention are findings from particular groups of people, not settled universal laws about everyone.
What confirmation bias actually is
We like to imagine we gather evidence like a fair judge, weigh both sides, and arrive at a conclusion. Confirmation bias is the well-documented tendency to do something closer to the opposite: to seek out, favor, and remember what fits what we already believe, while quietly undervaluing what doesn’t. Mostly we don’t notice it happening.
The cleanest early demonstration came from the psychologist Peter Wason, who in 1960 ran what’s now known as the 2-4-6 task. He gave people the number triple 2-4-6, told them it followed a hidden rule, and let them propose their own triples to figure out what the rule was. People would guess something like “even numbers going up by two,” then test only triples that fit their guess. They rarely tried a triple designed to prove themselves wrong. The actual rule was simply any three ascending numbers, which most of them never discovered, because they only ever looked for confirmation.
The cognitive psychologist Raymond Nickerson, who did a 1998 review , defines the bias as “the seeking or interpreting of evidence in ways that are partial to existing beliefs, expectations, or a hypothesis in hand.” Note the two halves: seeking and interpreting. It isn’t only about what we go looking for. It’s also about what we do with evidence once it’s in front of us.
Even the same evidence pushes us apart
You might expect that if you sat two people who disagree down with the exact same balanced evidence, they’d drift a little closer together. One study found close to the opposite.
In 1979, the Stanford psychologists Charles Lord, Lee Ross, and Mark Lepper recruited 48 undergraduates, half who supported capital punishment and half who opposed it, and had them read two fabricated but carefully balanced studies, one suggesting the death penalty deters crime and one suggesting it doesn’t. Same evidence for everyone. Both sides came away rating the study that agreed with them as better conducted and more convincing, and reported their original views growing stronger. The authors summed up their finding plainly: the result of exposing contending factions to identical evidence may be “not a narrowing of disagreement but rather an increase in polarization.”
How does that happen? People who hold strong opinions on a charged issue, the authors found, are apt to “accept ‘confirming’ evidence at face value while subjecting ‘disconfirming’ evidence to critical evaluation.” The study that agrees with you gets a nod. The one that doesn’t gets cross-examined for flaws. And memory does its own filtering on top: the same authors proposed that the bias may involve a propensity “to remember the strengths of confirming evidence but the weaknesses of disconfirming evidence.” Weeks later, what you carry forward is a lopsided file.
Why this isn’t a sign you’re foolish
It would be easy to read all this as a catalog of human foolishness. I don’t think that’s right. The same Stanford authors framed biased interpretation as something closer to a feature than a bug, arguing that willingness “to interpret new evidence in the light of past knowledge and experience is essential for any organism to make sense of, and respond adaptively to, its environment.” We can’t reweigh everything from scratch every time. We’d never get out the door. The bias may well be a shortcut for handling more information than we could ever impartially process, and it runs hardest on the beliefs we care about most.
What you can actually do about it
You’re not going to turn yourself into a perfectly rational machine, and I’d be suspicious of anyone selling that. But there are a few honest moves.
The most direct one is to do what Wason’s participants wouldn’t: go looking for the case that would prove you wrong. Deliberately steelman the other side, build the strongest version of the argument you disagree with, the one a smart person who holds it would actually make. If you can’t, you probably understand your own position less well than you think.
The other move is subtler. Notice the warm, satisfying click of agreement when a piece of evidence lands neatly on top of what you already believed. That feeling is real and pleasant, but it isn’t a verdict. Treat it as a flag rather than a confirmation. The moment evidence feels especially gratifying is the moment to slow down and ask what you’d be saying about it if it had pointed the other way.
I don’t think we ever really beat this. I certainly haven’t. Perhaps, we just learn, slowly, to distrust the click.
None of this is heavy territory, but if you find yourself in a genuinely fraught disagreement, with family or otherwise, a good counselor or mediator is better placed to help than any blog post.
