Psychology research · Reference
What is confirmation bias?
Confirmation bias is the tendency to search for, interpret, favour, and recall information in a way that confirms what we already believe, while giving less weight to evidence that contradicts those beliefs.
Definition
Confirmation bias describes a family of related habits: preferentially seeking evidence that supports a belief, interpreting ambiguous evidence as favourable to it, and more readily remembering supportive cases. The term was popularised by the psychologist Peter Wason, whose 1960 rule-discovery experiments showed that people tend to test hypotheses by looking for confirming instances rather than trying to falsify them. Because the bias shapes which evidence is collected and how it is read, it can leave a belief feeling well supported even when the underlying data are weak or mixed.
How it works
The bias acts at several stages of reasoning. At the search stage, people ask questions and select sources likely to yield agreement. At the interpretation stage, the same ambiguous result is read as supportive by a believer and dismissed by a sceptic. At the memory stage, confirming episodes are recalled more easily than disconfirming ones.
Crucially, confirmation bias is not the same as deliberate dishonesty. It is an unintentional, motivated pattern of attention and recall that even careful, well-trained people display. This is why simply intending to be objective is not a reliable defence against it.
Relevance to research
In research, confirmation bias threatens objectivity at every step, from designing a study to interpreting results. An investigator who expects a particular outcome may unconsciously favour analyses that produce it. This is a central reason methodologists insist on safeguards such as blinding, where those collecting or assessing data do not know group assignments, and pre-registration, where hypotheses and analysis plans are recorded before data are seen. These procedures remove much of the discretion through which confirmation bias would otherwise operate.
Significance for methods
Confirmation bias is one of the most studied cognitive biases and a key motivation for the open-science reforms that emphasise transparency and replication. Techniques such as considering the opposite, actively seeking disconfirming evidence, peer review, and adversarial collaboration are designed to counteract it. Recognising the bias also underpins good practice in literature reviews and meta-analysis, where systematic, pre-specified search strategies guard against cherry-picking supportive studies.
Key facts
At a glance
- Type: cognitive bias affecting search, interpretation, and recall
- Core tendency: favouring information that confirms prior beliefs
- Popularised by: Peter Wason's hypothesis-testing experiments (1960s)
- Operates: largely unconsciously, not as deliberate dishonesty
- Research safeguards: blinding, pre-registration, adversarial review
- Related to: belief perseverance and motivated reasoning
Common questions
FAQ
What is a simple example of confirmation bias?+
Someone who believes a particular diet works may notice and remember every time they feel better after eating it, while overlooking the times it made no difference. The supportive cases stand out and the contradicting ones fade, making the belief feel confirmed.
How does confirmation bias affect research?+
It can lead investigators to design studies, analyse data, or read ambiguous results in ways that favour their hypothesis. Safeguards such as blinding, pre-registration, and peer review are used precisely to limit this influence.
Can confirmation bias be eliminated?+
It cannot be fully eliminated, because it operates unconsciously. It can be reduced by deliberately seeking disconfirming evidence, considering alternative explanations, and using structured methods that remove discretion from data collection and analysis.
The step most authors miss
Doing CRediT right? Don’t stop at the statement.
A CRediT statement credits you inside one paper. The recognition CRediT was built for happens when those roles are tied to you, persistently. Sign in with your ORCID — free — and claim your CRediT contributions on casrai.org, the home of the standard. They become a verified, portable part of your identity, not a line that disappears into one PDF.
Free: claim your contributions, then export a journal-ready CRediT statement, schema.org structured data, JATS XML, CSV or BibTeX — and preview your public profile. A membership publishes that profile publicly and verifies the journals you serve.
