Openness, intelligence, and self-report intelligence

Past studies that have examined the relationship between Openness and crystallized ability have failed to account statistically for the fact that subtests commonly regarded as measures of crystallized intelligence (e.g., Vocabulary) are contaminated substantially by general intelligence. A method using residuals derived from a regression is proposed as a means to estimate crystallized ability. Further, self-reported intelligence was hypothesized to moderate or mediate the correlation between Openness and crystallized intelligence. It was found that two factors (General and Objective) could be derived from a principal components analysis (PCA) of the six Openness facets and that only Objective Openness correlated with intelligence. Using the residual approach to estimating crystallized ability, it was found that Objective Openness correlated with only g and not crystallized ability. The self-report intelligence measure correlated only with the General Openness factor. Based on the results of this study and a review of the empirical literature, it is argued that Openness should be correlated with g and that this correlation should exist based on theory. It is recommended that future research use the residual approach to estimating crystallized ability, and that a new emphasis on understanding the relationship between Openness and g is required. (C) 2004 Elsevier Inc. All rights reserved.