Metamemory and aging: Relations between predicted, actual and perceived memory task performance

Abstract
Four experiments examined adult age differences in predictions and postdictions of memory task performance. the results support the conceptualization of performance predictions as constructed judgments that are influenced by a number of factors, including memory self-efficacy and task appraisal processes. Prediction accuracy varied as a function of the type of task (recall better than recognition), subject age (better accuracy by old adults), and task experience (improvement over trials). Prediction accuracy appeared to be influenced by inferences about possible levels of task performance. Different age groups were equally accurate at postdicting performance. Three of the four experiments obtained predictions both before and after study, observing an increase in the correlations of predictions with recall after study. This upgrading effect was more pronounced for younger adults relative to older adults, possibly indicating poorer monitoring of learning or the contents of memory by older adults during the construction of the after-study prediction.