Factor analysis is a statistical procedure that identifies clusters of related items or variables. It reduces a large set of measurements to a smaller number of underlying dimensions, or factors.

In intelligence research, factor analysis of many different mental tests revealed a general factor, g, that accounts for much of the shared variance, along with more specific factors for verbal, spatial, and other abilities. In personality research the same method produced the five-factor model—openness, conscientiousness, extraversion, agreeableness, and neuroticism—by showing that thousands of trait adjectives cluster into these broad dimensions.

The technique examines patterns of correlation. Variables that correlate strongly with one another are assumed to reflect a common underlying factor. Different extraction and rotation methods can yield somewhat different solutions, so interpretation requires both statistical judgment and theoretical sense. Factor analysis is exploratory when the structure is unknown and confirmatory when a hypothesized structure is tested.

When a mass of overlapping measures needs to be simplified into coherent dimensions, factor analysis is the standard tool.