Types of cognitive biases

Confirmation bias

This bias leads us to seek out, interpret and favour information that confirms our existing beliefs, while ignoring or downplaying contradictory evidence. One study showed that individuals tend to overestimate the correlation between two variables when they correspond to their pre-existing beliefs, thereby biasing their interpretation of the data. arXiv

Availability bias

This bias leads us to estimate the probability of an event based on how easily similar examples come to mind. For example, one study showed that individuals overestimate the frequency of events covered in the media, such as plane crashes, because of their high media presence, even though they are statistically rare.

Representativeness bias

This bias leads us to draw conclusions based on stereotypes or pre-existing mental patterns rather than on objective evidence. An experiment revealed that participants judged the likelihood of a stereotypically described person being an engineer or lawyer based more on the description than on actual statistics on the distribution of these professions in the population.

Anchoring bias

This bias occurs when we rely excessively on initial information (anchor) when assessing a situation, even if it is incorrect or irrelevant. Research has shown that numerical estimates can be influenced by arbitrary numbers presented beforehand, with individuals insufficiently adjusting their judgement based on this initial anchor.

Negativity bias

This bias leads us to place more importance on negative information than positive information. One study found that individuals give more weight to negative criticism than positive praise, which can affect their perception of others and their decision-making.

Overconfidence bias

This bias causes us to overestimate our own competence, knowledge, or the accuracy of our predictions. The Dunning-Kruger effect illustrates that individuals who are least competent in a field tend to overestimate their abilities, while those who are most competent often underestimate theirs. Find out more...

Conformity bias

This bias causes us to change our opinions or behaviours to match those of the majority group, even if it contradicts our initial beliefs. Asch's experiment showed that 75% of participants conformed their incorrect answer to that of the majority at least once, despite clear evidence to the contrary.

Recency bias

This bias leads us to place more weight on recent information than on older information. For example, in performance evaluations, managers may be disproportionately influenced by an employee's recent actions, neglecting their past performance.

Projection bias

This bias leads us to believe that others share our own beliefs, attitudes, and experiences. One study found that individuals overestimate the extent to which their opinions are shared by others, which can lead to misunderstandings and misinterpretations.

Equiprobability bias

This bias consists of assuming that all events have an equal probability, even when this is not the case. For example, when throwing two dice, individuals often think that each possible sum is equally likely, whereas some sums have a higher probability due to the number of possible combinations to obtain them.

Hindsight bias

Also known as the «I knew it all along effect,» this bias leads us to perceive past events as being more predictable than they actually were. A meta-analysis of 128 studies confirmed the robustness of this bias, showing that individuals tend to overestimate their ability to predict an event after it has occurred.

Abusive generalisation

This bias leads us to draw general conclusions from a limited or unrepresentative sample. For example, in public health controversies, hasty generalisations may be used to minimise or exaggerate the risks associated with certain behaviours or products, based on isolated examples rather than solid statistical data. Find out more...