Short-term memory, working memory, and verbal abilities: How do they relate?

Abstract Subjects performance on short-term memory (STM) spans, STM probe-recall tasks, and complex working memory (WM) spans was used to assess the relationship between STM and WM, and to test whether these measures independently relate to verbal abilities. Factor analysis indicated that scores on the STM spans and probe-recall tasks loaded on a factor that was distinct from the WM spans, and regression and part correlations showed that these different factors accounted for separate variance in the Verbal Scholastic Aptitude Test (VSAT). These results provided evidence that STM and WM are different cognitive constructs, both of which are important to verbal abilities. It was also shown that within STM measures, rehearsal can obscure the relationship between STM capacity and abilities. For example, in the probe-recall tasks, only performance on final list items correlated with VSAT, and it was argued that these items were the most recently represented in STM, but had the least opportunity to be rehearsed. It was also suggested that digits are easier to rehearse than words, and we found that a STM word span correlated with VSAT, but a STM digit span did not. Furthermore, in considering each serial position in the probe-recall tasks, fewer items correlated with verbal abilities when digits were used as stimuli than when words were used. These results converge on the notion that rehearsal drives down the correlation between STM capacity and abilities. Overall, it was concluded that STM and WM tasks do reflect diffferent cognitive constructs,both of which seem to be important in abilities. In addition, when predicting verbal abilities from a STM tasks, the best measure is one in which rehearsal is not strongly influencing performance.