Judgments of Knowing: The Influence of Retrieval Practice

Four groups of college students first learned two paired-associate lists. Two experimental groups were asked to predict the likelihood that the response term of each pair would be recalled on the final test trial; learning either involved alternating study and test trials or only study trials before the prediction task. For the two control groups, the prediction task was omitted. All four groups then learned a third list without test trials and made judgments of knowing for items in this list. Subjects who had received test trials showed consistently higher prediction accuracy on the first two lists but their accuracy decreased on the third list when test trials were absent. A major source of facilitation produced by the opportunity for retrieval practice was traced to subjects' knowledge of their previous recall performance.