Lecture 3: Cognitive Components and Psychological Processes

Abstract

As I emphasised in the title of an article I published some years ago (Carroll, 1976), psychometric tests studied in factor analysis are in many cases examples of tasks studied in cognitive psychology. In that article, I examined the tests in the “kit” of factor-reference tests that had been assembled by French, Ekstrom, and Price (1963), interpreting them systematically in terms of putative cognitive processes, cognitive operations and strategies, and different kinds of memory stores. The variety of possible “elementary cognitive operations” seemed quite small and finite. Most cognitive operations and strategies appeared to be different combinations and sequences of very basic processes such as “addressing sensory buffers”, “searching memory” or “manipulating cognitive representations of stimuli.” Further, the variety of individual difference factors appeared to result from the interaction of a small number of elementary processes with different stimulus classes and response modes, different sensory modalities and with different kinds of memory stores. A monograph issued in 1980 (Carroll, 1980) was an attempt to review a variety of individual difference relations in psychometric and experimental cognitive tasks, that is, studies concerned with the possibility of measuring important dimensions of human cognitive ability through various types of simple cognitive tasks. In the course of my review, it seemed necessary to develop a theoretical model of what I call “elementary cognitive tasks” (ECTs), and I devised a scheme for classifying ECTs and for analysing various experimental paradigms in terms of ECTs. I also reviewed 55 studies relevant to individual differences in ECTs, and examined and in many cases reanalysed through factor analysis 25 pertinent datasets in the literature. I concluded that promising dimensions of individual differences could be found in a number of domains, including basic perceptual processes, reaction and movement times, mental comparison and recognition tasks, retrieval and production of names and other responses from semantic memory, episodic memory tasks, and analogical reasoning and algorithmic manipulation tasks. Individual differences were found in both speed and accuracy dimensions of these tasks; generally, speed and accuracy were found to be unrelated, or to have low intercorrelations. Considerable evidence was found for relations of ECT performances with scores on conventional psychometric tests, but the nature of these relations was seldom clear, chiefly because the components of psychometric test scores had not been adequately identified. Despite all the methodological and other problems that presented themselves in this field, I concluded that the study of individual differences in ECT performances would be a profitable field for further research. Nevertheless I was not optimistic that tests of ECTs could replace more conventional psychometric tests in practical testing situations. I felt that the study of ECTs was more important for theoretical understanding of cognitive ability than for the development of practical testing procedures.