First-Order and Higher-Order Creative Ability Factors in Structure-of-Intellect Measures Administered to Sixth-Grade Children

Abstract

This investigation is one in a series of re-analyses of those correlation matrixes consisting of structure-of-intellect (SOI) creativity tests devised by the late J. P. Guilford and his associates. Application of more recent exploratory and confirmatory factor analytic methodologies in comparison with those employed by Guilford and his associates provides a means for obtaining more objective, parsimonious, and possibly insightful interpretations of constructs underlying his creativity measures. Thus, the twofold purpose of this study was to ascertain for a sample of 403 sixth-grade children evenly divided by gender (a) the extent to which the first-order factors identified by Merrifield, Guilford, and Gershon in Report 27 from the Psychological Laboratory at the University of Southern California (May, 1963) could be replicated and (b) the degree to which higher-order factors could account for the covariance among the SOI measures in the correlation matrix. Three of Guilford's hypothesized creativity factors of ideational fluency, sensitivity to problems, and expressional fluency were replicated accompanied by a fusion of two of his hypothesized factors of originality and spontaneous flexibility into one dimension. Confirmatory factor analyses indicated that a general dimension reflecting primarily divergent production of semantic content could account for a large portion of the identifiable covariance among creativity measures. Application of goodness-of-fit indexes revealed that oblique higher-order factor models associated with three types of operations, two kinds of contents, and six forms of products did account for slightly more covariance than did the general factor model. Among all substantive models the oblique first-order nine-factor model provided the optimal fit. It was concluded that Guilford and his associates tended to extract and to rotate too many factors and that a more parsimonious interpretation of the factor structure of creativity tests than that afforded by Guilford could be realized.