Anchoring Bias
Anchoring Bias
Anchoring bias is a cognitive bias in which a person relies too heavily on the first piece of information they receive — the anchor — when making subsequent judgements or decisions. Even when the anchor is arbitrary or irrelevant, it pulls estimates and decisions toward it. This happens because the brain tends to use that first number or idea as a starting point and adjusts insufficiently from it.
Anchoring operates primarily through System 1 thinking — the fast, intuitive mode of thought described by Kahneman and Tversky's Dual Processing Theory. When people lack time, motivation, or information to think carefully, they default to their initial reference point. As a result, their final judgement often stays too close to that anchor.
Key Research: Tversky and Kahneman (1974)
Amos Tversky and Daniel Kahneman conducted a landmark experiment demonstrating anchoring bias in action. High school students were asked to estimate the value of a multiplication sequence. Group 1 was given 1 x 2 x 3 x 4 x 5 x 6 x 7 x 8, while Group 2 was given 8 x 7 x 6 x 5 x 4 x 3 x 2 x 1. Both sequences contain the same numbers, so both should yield the same answer — 40,320. However, because participants only had five seconds, they had to estimate. Group 1, who started with the small number 1 as their anchor, produced a median estimate of 512. Group 2, starting with 8, produced a median of 2,250. Neither group came close to the actual answer, and the direction of the error was clearly influenced by the anchor.
Strack and Mussweiler (1997) extended this work using an implausible anchoring scenario. Participants were asked whether Mahatma Gandhi died before or after age 9 (low anchor) or before or after age 140 (high anchor). The high anchor group estimated Gandhi's actual age of death at 66.7 years, compared to 50.1 years from the low anchor group. The actual answer is 78. These results confirmed that even clearly absurd anchors influence final estimates.
Anchoring in Everyday Life: Anchoring bias shapes decisions across many domains. In legal settings, Englich and Mussweiler (2001) found that judges were influenced by sentencing recommendations from prosecutors. In consumer behaviour, original prices displayed as crossed-out figures anchor shoppers to a higher reference point. In salary negotiations, the first number mentioned becomes the anchor around which all further discussion revolves.
Evaluation: Anchoring bias is well-supported across multiple methodologies, and its effects are robust. However, most studies use lab settings with university students, which limits generalisability. Individual differences such as numeracy, expertise, and motivation affect susceptibility to anchoring. Additionally, increasing awareness of anchoring does not necessarily eliminate it, suggesting that the bias operates at an automatic, pre-conscious level.