This study is concerned with whether the correlation between complex working memory spans and reading comprehension occurs because the complex spans reflect the capacity of a structural working memory that plays a causal role in comprehension or because a third factor, word knowledge, plays a causal role in both the span tasks and comprehension. If the latter hypothesis is correct, the correlation between word span and reading comprehension should be large when span is tested with low-frequency words but should not occur when span is tested with very familiar words. Ninety college students were tested on a simple and a complex version of the word span task with high-and low-frequency words. The Verbal Scholastic Aptitude Test (VSAT) was used as a measure of reading comprehension. The correlation between span and VSAT was somewhat higher when span was tested with low-frequency words, but was significant with both low-and high-frequency words. This suggests that both word knowledge and a content-free working memory play a causal role in the relationship between word span and higher level cognitive tasks. This research was supported by AFOSR 870069 to Randall W. Engle. We would like to thank Russel Onley for help in running subjects, Linda LaPointe for help with stimuli and analysis, and Lee Kirkpatrick for advice about the analysis.

No abstract