Overview

In psychology, similarity describes the subjective sense that two or more items, ideas, or events belong together or share important attributes. It is a core concept in cognitive psychology because judgments of similarity influence categorization, memory retrieval, learning, and reasoning. Rather than being a single, fixed relation, similarity depends on perceived properties, context, goals, and prior experience.

Major models and mental representations

Researchers have proposed several broad ways people mentally represent similarity. These ideas offer different accounts of why two objects seem alike:

  • Feature-based models: Similarity is computed from shared and differing attributes (common vs. distinctive features).
  • Prototype models: Categories are summarized by an ideal or average example; items are similar to the prototype by degree.
  • Exemplar models: Similarity is judged relative to stored examples; items close to many exemplars seem more similar.
  • Structural and relational models: Emphasis is on the alignment of relations and configurations rather than simple feature overlap.

Measurement and experimental methods

Psychologists measure similarity through several complementary techniques. Direct ratings ask participants to judge how similar two items feel on a scale. Reaction-time tasks infer similarity from speed and accuracy of classification or identification. Multidimensional scaling (MDS) and representational similarity analysis reconstruct perceptual or conceptual spaces from similarity data, revealing underlying dimensions. Experimental manipulations—such as changing context or goals—test how flexible similarity judgments are.

History and influential ideas

Interest in similarity stretches back to early categorization debates. Notable theoretical contributions include Shepard’s work on generalization and metric representations and Tversky’s contrast model, which emphasized asymmetries and the importance of distinctive features. Together, such accounts shifted thinking away from a single formal metric toward context-sensitive, descriptive models.

Applications and examples

Similarity affects many cognitive tasks: how children learn word meanings, how experts recognize patterns, how memory stores and retrieves items, and how people generate analogies. In applied settings, similarity principles guide clustering in machine learning, design of educational materials, and forensic matching. Practical examples include grouping animals by common traits, judging whether two melodies ‘‘sound the same,’’ or deciding which candidate best fits a job description.

Distinctions and notable facts

Important empirical observations include context dependence (the same pair can be more or less similar in different situations), asymmetry (A can seem more similar to B than B to A), and the role of goals and attention in shaping which features matter. Contemporary work often integrates multiple models and uses computational methods to predict when each account best explains human behavior. For further reading and overviews, see materials by contemporary researchers in the field.