Cognitive Abilities

Born or Made?

Cognitive abilities measured by conventional psychometric tests are traditionally seen as aptitudes or natural abilities (or even labeled “intelligence”), sharply distinguished from achievement tests. Now the consensus seems to be that there is at least a substantial overlap between the two kinds of measurements; that is, cognitive abilities so measured are developed or acquired, rather than innate. Scholars differ, however, as to whether the distinction between ability and achievement measures should be maintained. Some scholars view cognitive abilities as a form of developing competence, not that different from what is seen as achievement. Others argue for the distinct characteristics of cognitive abilities versus achievement, such as malleability (i.e., how easily they can be changed) in their development and generality (e.g., how widely they can be applied) in their functioning. The problem can be partially solved by looking at items in cognitive ability and achievement tests to see what kinds of knowledge base and cognitive processes these items tap into. The key issue is construct representation. To the extent that content knowledge is the focus of a test, it can be seen as by and large an achievement test; to the extent that a test demands the use of knowledge for reasoning and problem solving, cognitive abilities are involved. With regard to predictive validity, the best predictor of future achievement in a domain is the current achievement in that domain. Thus achievement can be treated as an important “aptitude” for future learning in that domain. The ability to reason in the symbol systems of a domain constitutes another important predictor of future success.

How Many?

Besides the ability–achievement distinction, there has been a perennial debate over how many cognitive abilities exist, and how broad or narrow these abilities are. Some researchers prefer parsimony; hence Charles Spearman's g or Raymond Cattell's broad distinction between fluid and crystallized intelligence. Others, from L. L. Thurston to Howard Gardner, prefer diversification of mental functions rather than some general, central capacity. John Carroll's three-stratum hierarchical model, which specifies three levels of cognitive abilities from the narrow to the broad, provides a compromise. A misconception that easily follows is that the hierarchical structure of cognitive abilities so defined reflects the structure of mind. In fact, it is merely a convenient classification system or taxonomy that helps organizing a wide range of tests rather than a psychological model. A true dilemma facing development of cognitive ability tests is the consideration of practical utility versus construct representation. The broader the range of tasks involved in a test (e.g., traditional IQ tests), the more practically useful, but the more obscured in its psychological meaning. Nevertheless, both interindividual differences in levels and intra-individual patterns of cognitive abilities are predictive of future achievement and developmental trajectories, suggesting the utility of using a more differentiated, multifaceted approach to ability measurement.

...