Journal · Evidence
Confirmation bias, and why you cannot self-diagnose it
Knowing about a bias does very little to protect you from it. People shown direct evidence of their own bias continue to deny it, sincerely.
Can you detect your own cognitive biases? Largely no. Research on the bias blind spot found that people readily identify bias in others while denying it in themselves, and they keep denying it when shown evidence that they have just displayed it. The finding has been replicated in a preregistered study.
The finding comes from Pronin, Lin and Ross in Personality and Social Psychology Bulletin, 2002, and a preregistered replication published in Collabra: Psychology in 2024. Participants recognised bias in others' judgements and rejected it in their own. When shown that they had just exhibited a specific bias, they explained why their case was different.
Two mechanisms were identified. People weight introspection heavily when assessing their own objectivity, and bias mostly does not show up in introspection because it operates below it. And people treat their own perception as a direct read of reality, which makes anyone who disagrees the biased party by definition.
They were not posturing. They genuinely believed their read was the unbiased one.
What this rules out
It rules out the questionnaire that asks which biases you have. Anything built on self-reported bias is measuring self-image, and self-image on this particular dimension is systematically wrong in a known direction. Any product that opens by asking you to rate your own blind spots has told you something about its rigour.
What works instead
- Observe the trade-offs someone actually makes on a real decision, rather than asking them to describe their tendencies.
- Fix beliefs in writing before the outcome, so hindsight cannot revise them.
- Externalise the objection. The counter-argument you did not generate is worth more than the one you did.
- Make the disconfirming search a step, not an attitude. "What would make this wrong" as a required field beats "stay open-minded" as an intention.
Five Peers editorial note. Sources named in the text.
We never ask which biases you have. We watch which trade-offs you make.
Compose my board