ABSTRACTThe work in this thesis is a development and extension of the work begun originally by Drs. Nathan Kogan and Karl Jöreskog. Kogan's work in creativity testing and its relationship to intelligence and achievement is the basis for the content of the thesis. Jöreskog's work in the development of factor analytic techniques is the basis for its methodology.In 1965, Wallach and Kogan (1965a) published the results of their research on creativity in the text Modes of Thinking in Young Children. They studied a group of children with a mean age of 10 years, 8 months using materials on creativity which have come to be referred to in the literature as the Wallach and Kogan tests of creativity. They provided evidence for the existence of a creativity dimension distinct from intelligence.Since 1965, William Ward (ETS), Cropley and Maslany (University of Saskatchewan), and Wallach and Wing (Duke University) have administered the Wallach and Kogan materials to elementary school children and to college students. They corroborated, with varying degrees of success, the Wallach and Kogan hypothesis.In this thesis, the original data of Wallach and Kogan, as well as the data from the Ward, Cropley and Maslany, and Wallach and Wing studies are reanalyzed using Jöreskog's Unrestricted Maximum Likelihood Factor Analysis (UMLFA). In addition, three of the Guilford studies and the Getzels and Jackson 1962 study on creativity and intelligence have been reanalyzed using the UMLFA technique. These reanalyses provide clear evidence not only for the distinct creativity and intelligence dimensions but also for figural and verbal subfactors in the creativity dimension. In some cases, the previous data were not factor analyzed. In these cases, the UMLFA technique provides clear evidence for subfactors. In cases where the data were previously factor analyzed, the UMLFA technique provides for a clearer interpretation of the data. The difference between this unrestricted maximum likelihood technique and other factor techniques is that this technique is developed on the basis of underlying normally distributed variables and the existence of a likelihood function that is minimized.In addition to these reanalyses, the Wallach and Kogan materials were used in an original research study with secondary school students from a large high school in Brooklyn, New York. The two‐factor structure of the creativity dimension was clearly verified. In addition, the creativity measures were related to many more common measures of intelligence than have been reported in studies thus far. Also, the creativity and intelligence measures were related to actual school grades in science, mathematics, English, and social studies. Three factors clearly related to creativity, intelligence, and school achievement were identified using the UMLFA technique. The figural and verbal subfactors in the creativity dimension were also clearly identified. The independence of the creativity and intelligence dimensions was verified, and a marginal relationship between creativity and school grades was indicated in the factor structure. To further verify this final indication, indices of creativity, intelligence, and school achievement were developed. The school achievement index was then used as the dependent variable in a multiple regression with creativity and intelligence as independent variables. The regression was carried out with each variable alone, with both jointly, and in stepwise progression with intelligence first and creativity second. The conclusions indicated in the factor structure were confirmed in each case.In summary, then, this thesis provides rather good evidence for the existence of a creativity dimension containing figural and verbal subfactors which is independent from intelligence and marginally related to school achievement. This evidence was obtained by reanalyzing the data from previous studies and by an original research study with high school students.