Confirmation bias is the tendency to seek out, interpret, remember and give more weight to information that confirms what you already believe, while ignoring or discounting evidence that contradicts it. It is one of the most common and most damaging cognitive biases in research because it quietly turns a genuine search for the truth into a search for validation. This guide gives you a precise definition of confirmation bias, explains why it happens, walks through clear examples and a worked case, and sets out practical, evidence-based methods you can use to reduce it in your own work.
Imagine you are writing a research paper on why “Homework is Helpful”. If you only search for studies that support your view and skip past the ones that do not, you are not actually researching — you are building an echo chamber. That is confirmation bias in action, and it is a specific type of cognitive bias. For the wider family of biases that distort studies, see our hub on research bias.
What Is Confirmation Bias?
Confirmation bias refers to the tendency of people to favour information that confirms their pre-existing beliefs, expectations or values, while overlooking, dismissing or under-weighting information that challenges them. It operates at every stage of thinking: how we search for evidence, how we interpret it, and how we later remember it. Crucially, it usually happens unconsciously — most people are convinced they are being perfectly objective even as the bias is steering them.
The term was popularised by the English psychologist Peter Wason in the 1960s. In his classic “2-4-6” task, participants were given a number sequence and asked to discover the rule behind it. The actual rule was simply “any three ascending numbers”, but most people fixed on a narrower hypothesis (such as “increasing by two”) and then only tested sequences they expected to confirm it, rather than trying sequences that could prove it wrong. The result was that they clung to an incorrect rule with great confidence — a textbook demonstration of confirmation bias.
Confirmation bias is pervasive and can influence decisions in many fields, including finance, medicine, law, politics and academia. It is the psychological “blindfold” that stops us from seeing the full picture, and in research it can quietly compromise the validity of an entire study.
Looking for help with your research?
Our subject-expert writers help students plan, structure and write balanced, bias-aware research across every discipline.
What Causes Confirmation Bias?
Confirmation bias is not a sign of stupidity or dishonesty — it is a by-product of how the human mind manages information efficiently. Several psychological mechanisms drive it:
- Cognitive ease. Processing information that fits our existing beliefs takes less mental effort than wrestling with evidence that contradicts them. The brain prefers the path of least resistance.
- Protecting our self-image and identity. Beliefs become part of who we are. Evidence that we were wrong feels threatening, so we defend our position instead of updating it.
- Desire for certainty. Ambiguity is uncomfortable. Locking onto a conclusion quickly relieves that discomfort, even when more evidence is needed.
- Emotional investment. The more we care about an outcome — a hypothesis, a political view, a financial position — the harder we work to defend it.
- Selective memory. We recall belief-consistent events more readily than belief-inconsistent ones, so our “evidence from experience” is itself skewed.
- Social and cultural reinforcement. Friends, colleagues and online communities often share our views, so the information environment around us already leans towards confirmation before we begin searching.
Understanding these causes matters because the fix is rarely “try harder to be objective”. Since the bias is automatic and feels exactly like ordinary, careful reasoning from the inside, the real solution is to build structured habits and procedures that force contradictory evidence onto the table — something we return to in the final section of this guide.
Types of Confirmation Bias
Confirmation bias shows up in several distinct forms. Recognising which type you are prone to makes it far easier to catch yourself in the act.
Selective Exposure / Seeking
Individuals tend to expose themselves primarily to information sources that align with their pre-existing beliefs. For instance, they might only watch news channels or read publications that share their political or social views. Example: A person who believes in the efficacy of a particular health remedy might only read articles and testimonials that support its benefits, while ignoring scientific studies that debunk its effectiveness.
Selective Perception
Even when presented with balanced information, individuals may only notice and accept details that confirm their beliefs while ignoring those that challenge them. Example: In a debate about climate change, an individual might only register the data points that support their stance, ignoring the comprehensive evidence that points the other way.
Selective Recall / Memory Bias
People tend to remember events or information that align with their beliefs more than events or information that challenge them. Example: A student who thinks a particular professor is unfair might remember the few times they received a lower grade but forget the numerous times they were marked generously.
Interpretative Bias
This involves interpreting ambiguous evidence as supportive of one’s existing beliefs. Example: If an employee believes a co-worker is out to get them, they might interpret any small oversight or mistake by that colleague as a deliberate slight, even if it was entirely unintentional.
The Confirmation Trap
This occurs when people fall into a loop of gathering information until they find evidence that confirms their beliefs, often ignoring a wealth of contradicting evidence. This can overlap with a bias for action — a preference for action over inaction — where someone rushes to find affirming information instead of patiently analysing all the evidence. Example: Someone researching the existence of extraterrestrial life might keep searching until they find a single questionable source that agrees with them, while overlooking numerous credible sources that offer counter-arguments.
Attitude Polarisation
When confronted with conflicting evidence, individuals may become more entrenched in their original beliefs rather than updating them — the so-called “backfire effect”. Example: In the face of studies showing the safety and effectiveness of vaccines, some sceptics become more convinced of their supposed dangers.
Confirmation Bias in Hypothesis Testing
Rather than testing an idea through objective means, individuals might seek out scenarios or experiments designed to confirm their hypothesis. Example: A researcher might design a study with a particular outcome in mind, ignoring alternative experimental designs that could challenge their hypothesis.
Biased Weighting of Evidence
Here, people give more weight to evidence that supports their beliefs while devaluing evidence to the contrary. Example: In a discussion about the effects of a new policy, an individual might emphasise a single positive anecdote over a comprehensive study showing mixed results.
| Type of confirmation bias | What it affects | Quick research example |
|---|---|---|
| Selective seeking | How you search for evidence | Only reading studies whose abstracts already agree with you |
| Selective perception | How you read evidence | Skimming past the “limitations” section of a contrary paper |
| Selective recall | What you remember | Citing the one supportive trial you remember, forgetting the others |
| Interpretative bias | How you read ambiguous results | Reading a non-significant result as “almost significant” |
| Biased weighting | How much each source counts | Trusting a blog that agrees over a peer-reviewed study that does not |
| Attitude polarisation | How you respond to challenge | Becoming more certain after reading a strong rebuttal |
A Worked Example of Confirmation Bias
The clearest way to understand the bias is to trace it through a single, concrete case from start to finish.
But notice what Sarah is doing. First, she selectively perceives: she registers her neighbour’s mood as proof, without asking whether he had a stressful day or received bad news — explanations unrelated to the moon. Second, she selectively recalls: on nights with no full moon she forgets the many times people acted out of character, only logging odd behaviour when it coincides with a full moon. Third, she falls into the confirmation trap: if she searches online, she stops as soon as she finds one article agreeing with her, ignoring the large body of research that finds no lunar effect on behaviour.
The result is a belief that feels strongly evidence-based to Sarah, but is actually built on selective attention and memory. Had she instead asked, “What would I expect to see if the full moon had no effect?” and deliberately counted calm full-moon nights and chaotic ordinary nights, her own data would most likely have overturned the belief. That single reframing — looking for disconfirming evidence — is the antidote to confirmation bias.
The same pattern scales up from Sarah’s kitchen window to a doctoral thesis. A student who is convinced their intervention works will tend to highlight the participants who improved, footnote the ones who did not, and quietly favour the analysis that reaches significance. None of this is deliberate fraud — it is confirmation bias operating exactly as it does for Sarah, just dressed in academic language. That is why the only dependable safeguard is to make disconfirmation a fixed step in your process rather than an afterthought.
Why Confirmation Bias Matters in Research
For students and researchers, confirmation bias is not just an interesting quirk of psychology — it directly threatens the quality of your work. It has significant implications across several domains:
- Decision making: Confirmation bias can lead to poor decisions, because people overlook important information that challenges their assumptions.
- Science and research: The scientific method requires objective hypothesis testing. Confirmation bias can jeopardise the integrity of your research findings, which is exactly why distinguishing between primary and secondary sources and applying a rigorous source evaluation method are so important.
- Validity and reliability: A study contaminated by confirmation bias may appear robust yet measure the researcher’s expectations rather than reality — see our guide to reliability and validity for how this plays out.
- Sociopolitical issues: People entrenched in their beliefs contribute to societal polarisation, making it harder to find common ground or reach compromise.
The effect is amplified in fields that depend heavily on interpretation and on emotionally charged topics — from health and education to the influence of social media on public opinion. In each, the temptation to read the data as supporting your prior position is strong, and the cost of doing so is a study that cannot be trusted.
“The first principle is that you must not fool yourself — and you are the easiest person to fool.” — Richard Feynman, Caltech commencement address, 1974
How to Avoid and Reduce Confirmation Bias
Because confirmation bias is automatic, willpower alone will not remove it. The reliable defences are procedural — habits and checks you build into your research so that disconfirming evidence cannot be quietly ignored. The following steps work in practice:
- Actively seek disconfirming evidence. For every claim, deliberately search for the strongest evidence against it. Use search terms such as “criticism of”, “limitations of” or “evidence against” alongside your topic.
- Try to falsify your hypothesis, not confirm it. Following Popper and Wason, ask, “What result would prove me wrong?” and design your method to test for it.
- Steel-man the opposing view. Write the strongest possible version of the argument you disagree with before you critique it. If you cannot state it fairly, you do not yet understand it.
- Use blind or pre-registered procedures. Pre-registering your hypothesis and analysis plan, and blinding data coding where possible, stops you (consciously or not) from shifting the goalposts once results arrive.
- Invite peer review and devil’s advocates. Ask a supervisor or colleague specifically to attack your reasoning. Diverse readers spot blind spots you cannot.
- Evaluate sources on quality, not agreement. Judge each source by its method and credibility, not by whether it supports your conclusion, and weigh contrary findings on the same scale as supportive ones.
- Keep an audit trail. Record every source you considered — including the ones you rejected and why — so selective omission becomes visible.
| Biased research habit | Bias-aware alternative |
|---|---|
| Searching only for support for your thesis | Searching equally for the strongest counter-evidence |
| Reading abstracts to confirm your view | Reading full methods and limitations of contrary papers |
| Designing a study to prove you are right | Designing a study that could prove you wrong |
| Citing the sources you remember | Logging every source, including rejected ones |
| Deciding the conclusion before analysis | Pre-registering the hypothesis and analysis plan |
Confirmation bias rarely travels alone. It interacts with related distortions — such as the vividness bias, where a single dramatic case outweighs dull statistics; omitted variable bias, where a hidden factor is ignored because it does not fit the expected story; and the Pygmalion effect, where expectations shape outcomes. Building the disconfirmation habits above protects against the whole family at once and keeps your conclusions honest.
Worried bias is creeping into your dissertation?
Our subject specialists help you design balanced methods, evaluate sources objectively and keep your findings credible from proposal to submission.
