Cognitive biases
Confirmation bias: why we notice what we already believe
“A coworker replied with one word again. Of course — they don’t like me.” This thought feels like an observation. Last week the same coworker helped you with a report, but that episode somehow never made it into the “evidence folder.”
This is confirmation bias: a tendency to look for, notice, interpret, and remember information in a way that supports an opinion we already hold. It is not stupidity or dishonesty. It is how ordinary thinking works once it already has a version of events.
Below, we look at how it shows up, why it happens, and what helps you test your beliefs instead of only confirming them. In the middle of the article there is a classic puzzle that makes it easy to catch the bias in yourself.
Contents
In short: how confirmation bias works
What confirmation bias is
The psychologist Raymond Nickerson, in a large review from 1998, called it a phenomenon “in many guises.” It turns up in everyday arguments, in medicine, in courts, in science, and in politics. And it shows up at three different stages.
What we look for. To test an opinion, we more often look for what will confirm it. A person who is sure that coffee is harmful types “dangers of coffee” into the search bar, not “benefits of coffee” or “coffee research.”
How we interpret. The same ambiguous detail gets read in favor of the version we already have. A coworker’s short reply means “they’re angry,” although it could mean “they’re busy.”
What we remember. Episodes that fit the version come to mind more easily. A month later you have ten examples of your coworker’s coldness in your head and not one example of their help.
Try it: the 2-4-6 task
In 1960, the psychologist Peter Wason devised a simple task. It is still the best way to show how this trap works.
Step 1. A secret rule was used to build sets of three numbers. All you know is that the triple 2, 4, 6 fits the rule.
Step 2. Work out what the rule might be and write down your guess.
Step 3. Test it: enter your own triples, and we answer “fits” or “doesn’t fit.” You can run as many tests as you like.
Step 4. When you are sure, reveal the rule and see how you tested.
Why numbers? The task has no emotions, no arguments, and no personal stakes. If the trap works even here, then it is not about stubbornness. It is about how we usually test our guesses.
What usually happens. A person quickly comes up with a guess such as “each number is two more than the one before” and tests it with triples that fit it: 8, 10, 12 or 20, 22, 24. Every time they hear “fits,” and their confidence grows. In Wason’s study, many participants named a rule that was narrower than the real one.
Why it did not work. All of these triples would fit both the real rule and the guess. So none of them could show that the guess was wrong. To find the mistake, you need a triple that your guess says should get “doesn’t fit,” for example 1, 2, 3. If it turns out to fit, the guess collapses, and you have learned something new.
Why it happens
Imagine you already have a ready-made folder in your head labeled “they don’t like me.” It is easier to drop each new episode into an existing folder than to start a new one. This is an image, not a description of a mechanism. But it conveys the main point: a ready-made version saves effort, and we gladly use it.
There is good news too. Testing a hypothesis on the cases that fit it is often a sensible strategy. Joshua Klayman and Young-Won Ha showed that in many real-life situations it gives useful information. The problem is that in some tasks, like 2-4-6, such tests cannot show that the hypothesis is wrong.
Then motivation joins in. A meta-analysis of studies on how people select information found that, on average, people more often choose information that agrees with their views. The effect is moderate, and it is stronger when the belief matters to the person. Choosing what feels pleasant is easier for us than choosing what is accurate.
And even when the facts in front of us are the same, we read them differently. In a classic experiment, supporters and opponents of the death penalty read the same two studies with opposite conclusions. Each side found the one that matched its position more convincing, and looked for methodological flaws in the other.
Where it gets in the way
Relationships. Once a couple has settled on a belief like “they’re always late” or “they don’t care about me anyway,” every new episode becomes one more piece of proof. The five times the person arrived on time go unnoticed. The sixth time, when they were late, goes into the folder. The argument quickly turns into an exchange of dossiers, where each side has its own collection of facts.
Work and decisions. When a decision has already been made, the data to check it is often collected in a way that makes it come out confirmed. In a job interview, the first impression of a candidate affects which questions they are asked next. Inconvenient numbers are explained away by “special circumstances,” and convenient ones are accepted at once.
News and the feed. We choose sources that feel good to us, and the feed adapts to what we read. As a result, it is easy to end up where everyone around us “obviously” thinks the same. This resembles the anchoring effect: the first version sets a frame, and the search then takes place inside it.
How you see yourself. The belief “nothing ever works out for me” also works like a folder. Mistakes are remembered, and successes are put down to luck. The longer this goes on, the more convincing the folder looks.
How to test your beliefs
The advice “be objective” is almost useless. In an experiment by Lord and colleagues, asking people to be objective changed little, while a concrete instruction — “consider what conclusions you would have reached if the data had been the opposite” — reduced bias noticeably. What works is a procedure, not an intention.
Confirmation bias is one of the biases on the cognitive bias map, where you can filter them by area of life.
When to see a psychologist
Ordinary confirmation bias is something everyone has, and by itself it is not a reason for a consultation. A reason appears when one belief about yourself or someone close is “confirmed” in every episode and starts to run your life: “nobody values me,” “they’re cheating on me,” “I’ll ruin everything.”
If such a thought keeps you in constant tension, damages relationships, or gets in the way of work, it is worth working through with a specialist. Cognitive behavioral therapy has techniques that help test such beliefs against facts rather than feelings. In a consultation, you can look at what your “folder” rests on and what it leaves out. You can book a consultation online.
Frequently asked questions
Is confirmation bias the same as stubbornness?
No. Stubbornness is a conscious unwillingness to change your position. Confirmation bias is usually invisible from the inside: the person sincerely believes they are simply looking at the facts.
Why do I easily see it in others but not in myself?
This is a known effect in itself. In a study by Emily Pronin and colleagues, people considered others more susceptible to biases than themselves. That is why it is more useful to rely on testing procedures than on the feeling of your own objectivity.
Can I get rid of confirmation bias completely?
Probably not, and you do not need to. Ready-made versions save effort in most ordinary situations. The goal is more modest: to notice when the stakes are high, and to test the belief on purpose at those moments.
When does confirmation bias matter most?
When the stakes are high and the belief has become important to you: about a close person, about yourself, or about a decision you have already made. At those moments it is worth checking the belief on purpose, rather than only collecting evidence for it.
