ERPs during study as a function of subsequent direct and indirect memory testing in young and old adults.

Event-related potentials (ERPs) were recorded from young and older adults while words were studied during structural and semantic encoding tasks. Items were presented twice to assess repetition effects. Subsequent memory effects (i.e. Dm or difference in subsequent memory) associated with non-target study items were also evaluated. Memory for non-target study items was tested either indirectly (word stem completion) or directly (cued recall). There were small, but unreliable age differences (favoring the young) on both the indirect and direct tests. These small differences were consistent with previous results for stem completion performance, but were counter to expectation for the cued recall test, where young adults were expected to show clear superiority. We conclude, based on task considerations, that for cued recall, subjects may have adopted an 'implicit' retrieval strategy. Because older adults typically have little difficulty with implicit retrieval, they fared almost as well as the young on cued recall. Dm effects were reliable for the young only. As Dm is thought to reflect elaborative encoding processes, the larger Dm magnitudes in the young than the old suggest that the small, though unreliable, age-related performance differences that resulted may have been mediated by such elaborative processing on the part of the younger adults.