SIMPLE AND COMPLEX WORD SPANS AS MEASURES OF WORKING MEMORY CAPACITY

Five experiments investigated the effects of word length in simple word span tasks and complex operation and reading span tasks and the relationship between these tasks and reading comprehension. The first 2 experiments showed word length effects using both simple and complex memory span tasks and that both simple and complex span tasks correlated with reading comprehension. In the third experiment, articulatory suppression did not eliminate word length effects. The final experiments showed that articulatory suppression eliminated the effect of word length when words were sampled with replacement from small fixed pools but not when sampled without replacement from a large pool. The word pool effects were not a result of concreteness of the words. We conclude that the reading span does not measure a working memory specific to reading. Further, in immediate memory experiments, repeating words from trial to trial may lead to a more limited coding than is used with nonrepeated words.