There's a thought experiment worth sitting with. Imagine you hold a firm belief about something - a person, a policy, a decision you've already made. Now imagine you encounter a mixed set of evidence: some that supports your view, some that contradicts it, and some that's genuinely ambiguous.
What do you do with each piece?
Most likely: you find reasons why the supporting evidence is solid, the contradictory evidence is flawed, and the ambiguous evidence leans your way if you look at it right. The conclusion you arrive at confirms what you started with.
This isn't obstinacy. It's confirmation bias - and it operates at every stage of how you interact with information.
What It Is
Confirmation bias is the tendency to search for, interpret, focus on, and remember information in a way that confirms one's preconceptions while ignoring or discounting contradictory evidence.
It isn't a single error but a cluster of interlocking tendencies that operate across how information is gathered, read, and recalled.
Three Stages Where It Operates
Selective exposure. Before encountering evidence at all, people tend to seek out sources, conversations, and environments that are likely to confirm existing views. News feeds, social circles, and information diets are filtered before any specific fact arrives.
Biased interpretation. When mixed or ambiguous evidence is encountered, people weight it toward the view they already hold. The same data point reads differently depending on what the reader already believes. Ambiguity is resolved in the confirming direction.
Memory bias. After the fact, confirming details are more likely to be recalled than contradicting ones. The accumulated memory of evidence becomes skewed toward the existing belief, creating a false sense of accumulated proof.
Together, these three stages create a loop: your beliefs shape what you find, how you read it, and what you remember - and the accumulated result is a feeling of comprehensive, well-founded certainty.
A Decision in Context
Sarah, a sales manager, is convinced that the recently adopted customer-relationship-management system will increase quarterly revenue. Whenever the team's numbers rise, she points to the CRM as the cause and celebrates the win. When figures dip, she attributes the drop to temporary market fluctuations or seasonal effects, never questioning the software's impact. Over time, she only recalls the months with growth and forgets the periods of stagnation, reinforcing her belief that the CRM is delivering the promised improvement.
Sarah isn't being dishonest with herself. The bias doesn't require intent. The interpretation at each step feels reasonable. What the pattern shows is how the three stages of confirmation bias - interpretation, attribution, and memory - can reinforce a conclusion that the data, in aggregate, doesn't actually support.
What the Research Shows
Individuals who exhibit stronger confirmation bias report higher confidence in their beliefs, even when objective accuracy is held constant. More certainty, not more accuracy. The bias amplifies conviction rather than calibration.
Selective exposure to ideologically congruent media is associated with increased affective polarization among partisans - not just different views, but stronger negative feelings toward those with opposing views. The belief becomes more entrenched; the people holding different beliefs become more threatening.
And importantly: instructing participants to consider the opposite, or to list reasons against their initial hypothesis, reduces confirmation bias in reasoning tasks. The bias responds to deliberate countermeasures - it isn't fixed.
The Common Misunderstanding
Confirmation bias is widely understood as "only looking for confirming information." But the seeking is only one stage. The bias also operates in how information is interpreted once found, and in which information is remembered afterward. Someone who genuinely reads sources across the political spectrum can still exhibit confirmation bias through how they weight and retain what they read.
A second misunderstanding: awareness eliminates the bias. Research consistently shows that the bias often operates before conscious evaluation begins - in what feels like obvious or natural interpretation. Knowing about it reduces its influence; it doesn't remove it.
See Confirmation Bias in everyday decisions
Pick a life context to see how this bias can show up outside the textbook.
Shared Hobby Overrides Technical Check Despite Bias-Mitigation Prompt
A hiring manager uses a 'consider the opposite' prompt but dismisses it due to rapport, leading to a biased hire.
Scenario
Alex, a team lead on a backend service for real-time analytics, interviews Jordan for a developer role. During the second technical interview, Jordan mentions enjoying the same board game Alex plays during lunch. Alex feels an instant rapport and notes it as a strength. Later, Jordan hesitates when asked to write a small script in Python 3.11, the language used daily for the service. Alex initially writes down three reasons why Jordan might need extra time to learn the language, following a 'consider the opposite' prompt he read about. However, because of the shared hobby, Alex dismisses those reasons as irrelevant and focuses on the positive connection. After the interview, Alex rates Jordan high on cultural fit and only slightly lowers the technical score. When a senior reviewer raises the skill gap, Alex insists the enthusiasm will bridge it and overlooks the concern. The team hires Jordan, and in the first month Jordan needs extra training on Python 3.11, confirming the worry Alex had dismissed.
Where The Bias Enters
Alex sought confirming evidence (shared hobby) and interpreted ambiguous technical answers as supportive of his initial positive impression, while forgetting or downplaying the contradictory information about missing Python skills, even after attempting a bias-mitigation prompt.
Decision Check
Before finalizing the hire, spend two minutes listing three pieces of evidence that would disprove your belief that the candidate can learn Python 3.11 quickly.
This scenario is illustrative. It explains the pattern and does not claim a documented public case.
Sources
- Nickerson, R. S. "Confirmation bias: A ubiquitous phenomenon in many guises." Review of General Psychology, 2(2), 175-220. (1998)
- Wikipedia contributors. "Confirmation bias." Wikipedia, The Free Encyclopedia. https://en.wikipedia.org/wiki/Confirmation_bias

