Recognition and recall with precuing and postcuing

There is extensive evidence showing that, with short lists of items, tests of item recognition show mostly recency effects but serial recall shows mar ked primacy effects. However, these studies are generally precued so subjec ts know in advance the type of test they will get. Four experiments are rep orted in which recognition and recall are tested under comparable condition s but the type of test is either precued or postcued. When precued, the sta ndard results are obtained. When postcued and recall and recognition tests are equiprobable, recall was essentially identical to that under the precue d condition but the recognition functions showed no serial-position effects . When postcued and with the probability of a recognition test 10-30 times more probable than recall, the recall functions showed a substantial loss o f primacy. The magnitude of these effects seem to depend in part on whether subjects are given time to develop strategies for recognition in isolation from serial recall. The theoretical implications are discussed. (C) 2000 A cademic Press.