Higher‐order factors of creativity within Guilford's structure‐of‐intellect model: A re‐analysis of a fifty‐three variable data base

Abstract
Both exploratory factor analyses (varimax and promax solutions) and confirmatory maximum likelihood factor analyses were used to re‐examine a correlation matrix of 53 tests from a battery administered to a sample of more than 400 Air Force officers. The data base originated in a report from the University of Southern California Aptitudes Research Project (Guilford, Wilson, & Christensen, 1952), which was intended to identify factors of creative thinking. The major objective of this study was to ascertain whether the covariation among the test variables that were conceptualized as first‐order factors within the structure‐of‐intellect model could be explained parsimoniously in terms of a number of higher‐order creative abilities. Application of a relatively objective oblique exploratory factor analytic technique (promax) afforded a replication of four of Guilford's creativity factors—two divergent production constructs of ideational fluency and word fluency, one construct representing sensitivity to problems, and another identified as redefinition or flexibility of closure typically involving transformations. Although substantial support was found for higher‐order factor models which distinguished among five types of psychological operations and three kinds of test content, statistical indicators of closeness‐of‐fit suggested that a mixed model of both first‐order and higher‐order factors was required to describe creativity thinking, perhaps within some form of hierarchical ordering. In addition to recognition of divergent production as a key component of creative endeavor, it appeared that a higher‐order convergent production factor involving primarily semantic and symbolic transformations constituted a dimension of potential importance to the creative thinking of mathematicians, scientists, engineers, and inventors. It was hypothesized that in creative thinking a variety of psychological operations within a dynamic interactive system is employed almost simultaneously in a forward and backward manner.