In a prior issue of the Journal of Intelligence, I argued that the most important scientific issue in intelligence research was to identify specific abilities with validity beyond g (i.e., variance common to mental tests) (Coyle, T.R. Predictive validity of non-g residuals of tests: More than g. Journal of Intelligence 2014, 2, 21–25.). In this Special Issue, I review my research on specific abilities related to non-g factors. The non-g factors include specific math and verbal abilities based on standardized tests (SAT, ACT, PSAT, Armed Services Vocational Aptitude Battery). I focus on two non-g factors: (a) non-g residuals, obtained after removing g from tests, and (b) ability tilt, defined as within-subject differences between math and verbal scores, yielding math tilt (math > verbal) and verbal tilt (verbal > math). In general, math residuals and tilt positively predict STEM criteria (college majors, jobs, GPAs) and negatively predict humanities criteria, whereas verbal residuals and tilt show the opposite pattern. The paper concludes with suggestions for future research, with a focus on theories of non-g factors (e.g., investment theories, Spearman’s Law of Diminishing Returns, Cognitive Differentiation-Integration Effort Model) and a magnification model of non-g factors.
A parameter linked with G-factors and the binding number
