Two-Process Models of Recognition Memory: Evidence for Recall-to-Reject?

Abstract

According to two-process accounts of recognition memory, a familiarity-based process is followed by a slower, more accurate, recall-like process. The dominant two-process account is the recall-to-reject account, in which this second process facilitates the rejection of similar foils. To evaluate the recall-to-reject account, we reanalyzed two experiments from Hintzman and Curran (1994) in which subjects made word recognition judgments at different response deadlines, and we conducted two new recognition experiments using pairs of similar pseudowords. The new analyses included modeling at both the group and individual subject levels. The results did not provide any distinctive evidence for recall-to-reject. In addition to discussing this two-process account, we describe a one-process account of recognition, in which the nature of similarity information varies across the course of judgment.