Age, processing speed, working memory, and fluid intelligence in children

A fundamental question that has interested psychologists for more than 100 years is whether individual differences in processing speed can account for a substantial portion of the individual differences in intelligent functioning. The current study continues in this long tradition and examines the most recent form of this issue. During the past decade, researchers have added working memory as a potential mediating variable in the relationship between speed and intelligence. The unique contribution of the current effort is to consider how the links between these three cognitive variables (processing speed, working memory, and intelligence) might be causally connected in the context of development. That is, the question addressed here is: To what extent are age-related differences in intelligence mediated by age-related differences in processing speed and working memory? To this end, a correlational study that included age, processing speed, working memory, and intelligence was conducted on a sample of 219 individuals ranging from age 7 to 19 years. Processing speed was assessed using a battery of four reaction time tasks, working memory was assessed using a battery of four memory span tasks, and intelligence (or more specifically, fluid ability) was assessed using the Raven's Progressive Matrices. Results of path and multiple regression analyses revealed that nearly half of the variability in performance in fluid intelligence could be accounted for by age, speed, and working memory. Furthermore, of this explained variance over three-fourths could be accounted for by age-related differences. The current results support a multiple-mediation interpretation of age-related differences in fluid intelligence. That is, consistent with the multiple-mediation hypothesis, over three-fourths of the age-related differences in fluid ability were found to be due to age-related differences in working memory, leaving less than one-fourth of the age-related differences in fluid ability to be due to other factors unspecified by the current developmental cascade model. More important, however, was the finding that nearly all of the age-related difference in working memory was attributable to age-related differences in processing speed. Additionally, the current developmental cascade model permitted the assessment of individual differences in speed on working memory and fluid intelligence. Consistent with Jensen's (1993) theory, individual differences in speed had a direct effect on working memory capacity which, in turn, was a direct determinant of individual differences in fluid ability.