ABSTRACT
Resource models provide an alternative perspective on working memory to that of structurally oriented multiple component models. The general idea of resource approaches is that working memory draws on limited pools of energy-like resources (typically “activation”) and storage and processing compete for such resources. This approach tends to use individual difference measures of working memory span which involve both storage and processing; e.g., the “reading span” measure in which participants have to read a number of sentences aloud for meaning (processing) and then recall the final words of each sentence (storage). The maximum number of sentences for which final words are recalled is the span measure. It may be noted that reading span measures do not correlate strongly with simple word span for lists of unrelated words (which may be taken to measure short-term verbal memory), but rather reflect the ability to retain earlier material in the face of subsequent unrelated processing. Just and Carpenter (1992) applied this approach in the domain of reading comprehension. They found that individual difference measures of reading span, taken to reflect working memory capacity, predicted a number of aspects of verbal comprehension, particularly when the comprehension task was demanding.
