Crystallized Intelligence

Psychometric Evidence

The psychometric or statistical evidence in favor of the validity of gc as distinct from g rests largely upon empirical results that suggested the inadequacy of Spearman's tetrad equation for the verification of g. In simple terms, several researchers noted that the development of several spatially based tests of cognitive ability during the mid-1920s and early 1930s tended to correlate with others to a disproportionate degree than would have been otherwise predicted by the ircorrelation with verbally/education based tests (i.e., gc). If g were the only intelligence factor, the correlations would have been uniform across all subtests, within sampling variability. The plausibility of a gc-type factor, as distinct from g, has also been suggested based on more sophisticated analyses, such as Louis Thurstone's multiple factor analysis, which was designed to allow for the extraction of more than one factor from a correlation matrix. More recently, the observation of [Page 214]a gc factor, as distinct from g or gf, has also been supported based on confirmatory factor analyses.

Physiological Evidence

Cattell contended that, although psychometric evidence had a place in supporting the plausibility of gc, the neurophysiologically related evidence should be regarded as more crucial to its establishment. Foremost, there was evidence to suggest that the trajectory of cognitive ability adult development in gc is unique among cognitive ability factors in that performance tends to increase during early adulthood (20–30) and remains stable until the age of approximately 65, after which some decrement in performance is observed. Cattell also hypothesized that individual differences in gc would be more greatly influenced by the environment, in comparison to other cognitive ability factors. This hypothesis has been supported by behavior genetic studies, which have suggested that approximately 45% of gc variance may be heritable, which can be contrasted to the 65% heritability estimate associated with gf. Relatedly, performance on gc-like tests tends to be relatively resistant to brain injury and disease. Consequently, gc-type tests are referred to as hold tests in the area of neuropsychology and frequently form the foundation for an estimate of premorbid IQ in many neuropsychological assessments. Although relatively resistant to brain damage, gc performance has been empirically implicated to be mediated by Broca's area, which is an area of the left hemisphere of the brain that does not appear to need to be intact to perform other cognitive ability tasks. Also, in many neuropsychological cases, recovery in performance on gc-type tests can be expected in those who have attained physical maturity. This last empirical observation supported Cattell's investment theory of intelligence, which posits that individual differences in gc are due to individual differences in the previous application of gf. For this reason, a positive correlation of approximately .50 is often observed between gc and gf. Thus, in combination with individual differences in gf capacity, individual differences in gc were considered to arise because of nonintellective factors such as motivation, socioeconomic status, education, and personality.

...