How an initial value shapes a decision

A cognitive bias is a systematic tendency in thinking that can pull judgments away from a balanced assessment of relevant evidence. It isn't just any mistake. A bias follows a pattern in how a person processes information.

With anchoring bias, an initial value or reference point has a disproportionate influence on a judgment. That starting point is the anchor: a piece of information used as a reference for a later judgment. The key word is disproportionately. An informative starting point isn't automatically a source of bias; the problem arises when someone gives it more weight than its relevance or reliability warrants.

Imagine two people estimating how long it will take to edit an audio recording. Before they see the same task description, one hears a relatively short suggested duration; the other hears a relatively long one. Even with identical task information, their estimates may stay closer to their respective starting points. The diagram shows how different anchors can pull judgments in different directions. This pattern is possible, not inevitable.

Its role in decision-making

An anchor can set the range of what initially feels plausible. As more information becomes available, a person may adjust away from it—but not far enough. Another possibility is that the anchor makes information consistent with it easier to bring to mind. In either case, the final judgment can stay tied to the starting point instead of being built afresh from the evidence.

Consider the editing example. The person given the shorter estimate might focus on the recording's simple structure and overlook the time needed to correct sound problems. If they underestimate the workload, that could affect whether they accept the task or how much time they allocate to it. The chain is straightforward: the initial reference influences an estimate, which then influences a decision.

Someone else may supply an anchor, or a person may generate one from their own previous experience. An anchor doesn't have to be accurate to exert an influence. Even recognising that a starting value is questionable does not guarantee that its effect disappears.

Not every decision is affected in the same way. Relevant knowledge and clear task information may reduce reliance on an anchor, as may independently developed estimates. One sensible safeguard is to form a judgment from the relevant evidence before considering someone else's suggested value. The aim is to give starting information an appropriate weight, not to ignore it all.

How an existing belief filters evidence

Confirmation bias is a cognitive bias in which people favour information that supports an existing belief or expectation over information that challenges it. It can shape the evidence they look for and how they interpret what they find. Supporting a belief isn't itself biased. The problem is using different standards to judge supporting and opposing information.

Suppose a student already believes that one audio-editing application is the best choice for a project. They look mainly for favourable reviews. Praise seems trustworthy to them, while reports of technical problems are dismissed as user error. The decision looks research-based, but the student has filtered that research through the conclusion they preferred from the start.

A reinforcing loop can develop. The existing belief draws attention toward supportive evidence, which increases confidence in the belief. That greater confidence then encourages more selective searching. The diagram shows how this happens and where challenging evidence may be excluded.

Its role in decision-making

Confirmation bias can operate through several routes:

• Selective searching: looking for evidence likely to support the preferred option while neglecting alternatives.
• Biased interpretation: reading ambiguous evidence as support for the existing belief.
• Uneven scrutiny: questioning the quality of opposing evidence more strongly than that of supportive evidence.

A person can become more confident through these processes without having better support for their decision. The student in the application example may commit to software that does not meet the project's requirements because they have discounted relevant evidence about its limitations.

Decisions can also become resistant to revision. Even when new information is available, it may have little effect if the person dismisses it simply because it conflicts with their expectations. That doesn't mean every rejected claim should have been accepted. Evidence may genuinely be weak; the bias lies in judging it differently depending on whether it agrees with the prior belief.

Distinguishing the two biases

Anchoring bias concerns the excessive influence of a starting reference point. Confirmation bias, by contrast, concerns the preferential treatment of evidence that supports an existing belief. Anchoring can occur without a person actively defending the anchor, while confirmation bias can occur without any initial numerical estimate.

To reduce confirmation bias, a decision-maker can deliberately seek credible evidence against their preferred option, then assess both sides using the same standards. The aim isn't to reverse every initial belief. It is to let the quality of the evidence guide the decision, rather than whether the evidence agrees with the belief.