Delaying students' metacognitive monitoring improves their accuracy in predicting their recognition performance.

Abstract

The accuracy of students' judgments of learning (JOLs) in predicting recognition versus recall was investigated in 4 experiments. Students studied paired associates and made JOLs, which occurred either immediately after an item had been studied or shortly after an item had been studied. Students then received tests of associative recognition or paired-associate recall. JOL accuracy was greater for delayed JOLs than immediate JOLs, and the accuracy of JOLs was lower in predicting recognition than recall. The latter finding occurred (a) regardless of whether students had anticipated a recall test or a recognition test when making JOLs and (b) regardless of whether JOLs had been cued by only the stimulus of an item or by the entire stimulus-response pair. Correct guessing was shown to contribute to the lower accuracy of students' predictions of recognition.