Reading speed and prose memory in older and younger adults.

A method is described for determining the minimum amount of time required to read a proposition in the context of a sentence. This threshold reading time was reliable, stable, and significantly longer for older adults than for younger adults. Prose memory was assessed for texts presented at varying speeds determined by the threshold reading time measure and by a whole-text reading time measure. Recall from these paced conditions was compared with recall from a self-paced condition. In general, age differences in recall tended to increase as the amount of time allowed for processing increased. The results of regression analyses suggest that although the basic speed of processing for a single sentence accounts for substantial variance in prose recall, processes that operate across sentence boundaries are also important and require further specification.