Decomposing the age-related deficit in free recall

The temporal relations among list items exert a powerful influence on episodic memory retrieval. For example, in free recall, participants exhibit marked recency and contiguity effects: temporal proximity of an item to the time of test and temporal proximity of an item to the justrecalled item both serve to promote retrieval. Previous research has focused on the serial position curve as a means of assessing the role of recency-sensitive processes in memory retrieval. We employ a decomposition method to the serial position curve, partitioning performance into recency-sensitive and contiguity-sensitive processes (e.g., M. W. Howard and M. J. Kahana, 1999). Two experiments examined the age-related deficit in free recall. Although recencysensitive processes, reflecting self-initiated recall, were indistinguishable for young and older adults, contiguity-sensitive processes, reflecting temporal association, were markedly impaired in older adults. This finding of a clear associative deficit in normal aging constitutes direct evidence for M. Naveh-Benjamin’s (2000) associative deficit hypothesis, and suggests that this associative deficit may be responsible for older adults’ well-known impairment in free recall. Although older adults also made many more intrusions than the young, this cannot provide an explanation for the associative deficit. In contrast, an associative deficit could cause more intrusions by means of a lowered recall criterion.