A Short List of Cognitive Biases

Cognitive biases are systematic deviations from rational judgment. All of us experience them to some degree, and it's useful to understand them in order to avoid or mitigate them. A short list of ones that I've found helpful in my personal and professional life.

• Confirmation bias is the tendency to favor information that confirms (and discount information that disconfirms) one's pre-existing beliefs.

• Sampling on the dependent variable is selecting cases on the basis of meeting a criteria and then use those cases as evidence for the criteria.

• Cognitive Dissonance: the mental stress or discomfort experienced by an individual who holds two or more contradictory beliefs, ideas, or values at the same time. You can recover from cognitive dissonance by (1) adjusting behaviors, (2) acquiring new information that outweighs the dissonant beliefs, or (3) reducing the importance of the new fact.

• Framing bias / Prospect theory: when people react to a particular choice in different ways depending on how it is represented, e.g., when people value losses more than valuing gains.

• Fundamental attribution error: the tendency to exaggerate the role of others’ internal characteristics vs. situational factors when explaining their behavior.

• Mere Exposure Effect: people tend to develop a preference for things merely because they are familiar with them (e.g., words, sounds, people).

• Recency, primacy, similar-to-me, default bias: We tend to favor things we recently experienced, experienced,  were most similar to us, or are the status quo.

• Observer-expectancy effect: when a researcher expects a given result and therefore unconsciously manipulates an experiment or misinterprets data in order to find it.

• Survivorship bias: concentrating on the people or things that "survived" some process and inadvertently overlooking those that did not because of their lack of visibility.

• Selection bias: happens when the members of a statistical sample are not chosen completely at random, which leads to the sample not being representative of the population.