Illusory correlations

Two common sources

Prior expectation is one source: people expect certain things to go together and notice the cases that fit. Distinctiveness is another: rare events that occur together stand out. Remembering the striking cases while overlooking the mundane ones can give a false sense of a pattern.

• Believing bad weather makes joint pain worse based on memorable days
• Clinicians once linking drawing features to diagnoses without valid evidence
• Associating a small group with unusual behavior

Illusory correlations and stereotypes

David Hamilton and Robert Gifford showed in 1976 that participants judged a minority group more negatively even when the ratio of good to bad behaviors was the same in both groups. Because the minority group and undesirable behavior were both relatively uncommon, their co-occurrence seemed especially distinctive. This offered a cognitive explanation for how stereotypes could form without any prejudice at the start.

Alternative explanations

The distinctiveness account has been challenged. Klaus Fiedler has shown how such effects can arise from the way information is sampled and aggregated: with fewer observations of a smaller group, people may be less certain about its true rate, without any special attention to rare pairings. Researchers still dispute how much weight these different mechanisms deserve. The basic phenomenon is well documented, though its strength depends on the task.