The metamemory approach to confidence: A test using semantic memory

The metamemory approach to memory confidence was extended and elaborated to deal with semantic memory tasks. The metamemory approach assumes that memory confidence is based on the products and processes of a completed memory task, as well as metamemory beliefs that individuals have about how their memory products and processes relate to memory accuracy. In two experiments participants were asked deceptive and nondeceptive questions involving geographical information. In both experiments, as predicted by the metamemory approach to memory confidence, there was a positive confidence/accuracy relationship for nondeceptive items and a negative relationship for deceptive items. Experiment 2 used items that constrained the memory strategies (e.g., hierarchical reasoning about spatial location) used by the participants. The results supported the hypothesis that the participants were aware of the levels of memory accuracy associated with the different strategies and used that information to generate their memory confidence judgments.

► The metamemory theory of confidence was extended to deal with semantic memory (knowledge about geography).
► The theory assumes individuals have metamemory beliefs about relation of memory processes to memory accuracy.
► The theory predicts a positive relation of confidence and accuracy for nondeceptive semantic memory items.
► The theory predicts a negative relation of confidence and accuracy for deceptive semantic memory items.
► Data across experiments support the theory with large effect sizes.