Developmental bends in children's memory monitoring - Evidence from a judgment-of-learning task

Two experiments were conducted to explore young children's memory monitorin g abilities on a judgment-of-learning (JOL) task. Recent research on adults ' JOLs has shown that predictions about subsequent recall for items that ha ve been recently studied have never been very accurate immediately after le arning but have been very accurate when judgments were delayed. One of the major goals of the present studies was to investigate whether the delayed-J OL effect could be observed in children of different ages. A secondary goal of the study was to compare individual-item JOLs with aggregate JOLs based on all items of a given list. If young children possess basic monitoring s kills, both their delayed JOLs and their aggregate judgments should be comp arably realistic. Our two experiments confirmed this assumption for all age groups involved (kindergartners, second and fourth graders). That is, JOLs were much more accurate when given after a delay of about 2 min than immed iately after study, and overconfidence was typically larger for item-by-ite m JOLs than for aggregate-item JOLs. In fact, the pattern of findings for t he older school children was very similar to that found for adults. Overall , these findings support the position that developmental trends in children 's procedural metamemory are not due to differences in basic monitoring ski lls but attributable to developmental changes concerning the interplay betw een monitoring and self-regulation activities. (C) 2001 Elsevier Science In c. All rights reserved.