Cognitive Biases · CB-01
You don't just prefer evidence that confirms what you already believe — you actively stop looking once you find it.
The tendency to search for, interpret, favor, and recall information in ways that confirm one's pre-existing beliefs, while giving disproportionately less attention to information that contradicts them. It operates at every stage: which evidence you seek out, how you interpret ambiguous evidence, and what you remember afterward.
Extensively documented in modern cognitive psychology starting with Peter Wason's 1960 rule-discovery experiments; the underlying tendency was noted far earlier, including by Francis Bacon in his 1620 Novum Organum.
The Mechanism
Wason's 2-4-6 task — people test their hypothesis, not the rule
Wason's participants weren't stupid — they were systematically testing the wrong thing. They searched for sequences that would confirm their hypothesis and almost never tried one designed to break it, which is exactly the mechanism confirmation bias describes: seeking confirming evidence feels like verification, but it isn't.
01 · IT OPERATES AT THE SEARCH STAGE, NOT JUST THE INTERPRETATION STAGE
The bias shapes what evidence you even encounter
Confirmation bias isn't only about misreading evidence you're given — it shapes which evidence you go looking for in the first place. Left unchecked, you simply never encounter much of the disconfirming evidence that exists, because you never searched for it.
02 · IT'S STRONGEST WHEN A BELIEF IS TIED TO IDENTITY OR PRIOR COMMITMENT
Motivated reasoning compounds the effect
The bias is measurably stronger when a belief is bound up with self-image, a public commitment, or a costly prior decision — precisely the situations where objectivity matters most, since the emotional stake in being right (rather than accurate) grows in exactly those cases.
03 · THE FIX IS DESIGNED DISCONFIRMATION-SEEKING, NOT WILLPOWER
You can't out-want your way past it — you need a process
Simply resolving to 'be more open-minded' rarely works, because the bias operates below conscious awareness. Effective countermeasures are structural: explicitly asking 'what evidence would change my mind,' assigning someone the formal role of arguing the other side, or seeking out sources you expect to disagree with before finalizing a judgment.
Where It Fails / Inversion
Where it fails / inversion
Actively distrusting all confirming evidence overcorrects into a different error — sometimes the evidence that confirms your view is simply the correct evidence, and reflexively discounting it in the name of 'balance' is its own bias. The goal is proportionate weighing of all available evidence, not systematically favoring disconfirmation any more than confirmation.
How To Use It
Worked example · running a genuine pre-mortem before committing to a strategy
Before finalizing an important decision, explicitly write down what evidence would prove you wrong, and then go look for exactly that evidence rather than more evidence supporting your existing view. Teams that build this step into their process — a genuine devil's advocate role, not a symbolic one — catch more flawed strategies before commitment than teams that only review supporting evidence.
How to use it
Before treating a belief as settled, ask explicitly: what would change my mind, and have I actually looked for it? If you can't name a specific disconfirming test you've already run, your confidence is resting on search bias, not verification.
See Also