Merely labeling objects into two categories makes perceivers exaggerate differences between the groups.

Evidence

One famous piece of evidence for the biasing impact of categorization on perception comes from a study by Tajfel and Wilkes (1963). Like Allport, Tajfel and Wilkes conceived categorization (and stereotyping) to result in the overestimation of intergroup differences and intragroup resemblances. Participants estimated the length of a series of lines that varied from each other by regular increments. When shorter lines (As) were systematically given a different label than longer lines (Bs), participants overestimated the differences between the categories. This basic categorization effect applies to estimates about attitudes, traits, or even physical values (e.g., Krueger & Clement, 1994) and points to the biasing influence of categorization. However, it seems most pronounced when the judgment is uncertain, such as when participants communicate their estimates in an unfamiliar measurement unit (Corneille, Klein, Lambert, & Judd, 2002). — John F. Dovidio, Peter Glick, Laurie A. Rudman, On the Nature of Prejudice, 1155

Categorization

• CANDY: The absurd minimalism of the experiment — line lengths, arbitrary As and Bs — producing systematic overestimation is a light, striking elevator fact.
• QUESTION: The boundary effect is strongest under judgmental uncertainty — worth chasing whether uncertainty is the mechanism behind all categorical accentuation.