Preschool Children's Judgments of Learning

Preschoolers’ ability to make judgments of learning (JOLs) was examined in three experiments in which they were taught proper names for various animals. In Experiment 1, when judgments were made immediately after a name had been taught, nearly every child predicted subsequent recall of every name. When judgments were made after a delay of a few minutes, fewer (14 of 29) showed this response tendency. The delayed JOLs of those who predicted at least one recall failure were still overconfident, however, and were not correlated with final recall. In Experiment 2, children received a second study trial with feedback, made yes-no JOLs after a delay, and then completed a forced-choice task. In the forced choice task, an animal whose name had been recalled previously was pitted against an animal whose name had not been recalled. Compared to Experiment 1, more children predicted at least one recall failure, and predictions were moderately accurate. In the forced choice task, well-learned items were readily distinguished from less well-learned ones, even by children whose yes-no JOLs had been inaccurate. Experiment 3 examined the effect of providing an additional retrieval attempt with feedback on delayed JOLs. Half of the children received a single study session and the others received an additional study session with feedback. The number who predicted at least one recall failure did not vary by group. However, children in the practice group showed less overconfidence than children in the no practice group. In the forced-choice task, both groups showed the same high level of performance as the children in Experiment 2.