Two sources of information in reconstructing event sequence.

Abstract

Reconstructing memory for sequences is a complex process, likely involving multiple sources of information. In 3 experiments, we examined the source(s) of information that might underlie the ability to accurately place an event within a temporal context. The task was to estimate, after studying each list, the temporal position of a single test word within that list. In the first 2 experiments, we demonstrated that memory for temporal location was better following semantic encoding than silent reading of the list, which in turn was better than orthographic encoding of the list. Although other measures of sequence retention have revealed impaired memory for order with greater item-level encoding, these experiments demonstrated that item-level encoding improved memory for temporal-location. A 3rd experiment extended these findings by measuring interitem associations in addition to item memory, demonstrating that memory for temporal location within a list was more closely related to item information than to interitem relational information. It is now clear that reconstructing an event sequence can involve at least 2 distinct sources of information-both item and relational encoding can play important roles, depending on the nature of the test for order. (PsycINFO Database Record