Daneman, M. & Carpenter, P. (1980). Individual differences in working memory and reading. Journal of Verbal Learning and Verbal Behavior, 19, 450-466.

Research Question: What is the relationship between reading comprehension and working memory?

Methods: Twenty undergraduate students were given three tests including a reading span test to measure their working memory span, a traditional word span test, and a reading comprehension test that asked questions about facts and pronominal references.  After reading the passage the subjects were asked two questions – one was about a fact in the passage and another was about the pronoun in the last sentence and its antecedent mentioned earlier in the passage.  The distance between the pronoun and the antecedent varied among passages. 

Findings: The traditional short term memory tests of digit and word span tasks do not tax the processing component of the working memory test. Studies using such measures do not find strong correlations with reading. If a short memory test was used with heavier demands on processing, poor readers would have less attention left to retain information. The reading span test was correlated significantly .90 with the pronominal reference test and .72 with the fact retrieval task. Its correlation with the global assessment reading comprehension (Verbal SAT) was also significant: r = .59. The word span task correlated non-significantly only .33, .37, and .35 with the reading tests respectively.

The results indicate that the reading span task is related to working memory capacity.

Instructional Implications: The good reader has more functional working memory capacity available for the demands of chunking. He is more likely to have more concepts and relations from preceding parts of the text still active in working memory. The good reader's chunks should be richer, and more coherent, and contain different information.