THE RELATION BETWEEN REMEMBERING AND KNOWING AS BASES FOR RECOGNITION - EFFECTS OF SIZE CONGRUENCY

In three recognition memory experiments, subjects studied a list of randomly generated geometric shapes, followed by a recognition test in which old items were either size congruent (same size at study and test) or size incongruent. In Experiment 1, the process dissociation procedure (Jacoby, 1991) showed that changing the size of the items led to a decrease in both recollection and familiarity. In Experiment 2, the remember/know procedure (Tulving, 1985) showed that recollection, as indexed by the proportion of ''remember'' responses, decreased with size incongruence, but familiarity, as indexed by the proportion of ''know'' responses, increased. The latter effect along with other problems with the remember/know procedure were found to arise because of the procedure's underlying assumption that recollection and familiarity are mutually exclusive. When an independence assumption was combined with the remember/know data (IRK), results agreed with those of the process dissociation procedure. In Experiment 3, receiver operating characteristics (ROCs) were examined using the remember/know procedure and showed that familiarity was well described by a signal detection process that was independent of recollection. (C) 1995 Academic Press, Inc.