Metamemory refers to our knowledge and awareness of our own memory content and processes. Metamemory experiences are the subjective awareness of our own memory. Metamemory experiences include the feelings that generates a judgment of learning, and it includes tip-of-the-tongue states. For metamemory to be useful to an individual, it must accurately predict memory performance. We measure this ability to predict performance in a number of ways. Global accuracy refers to a judgment about metamemory knowledge as a whole, whereas local accuracy refers to metamemory about a specific item. Absolute accuracy (also known as calibration) refers to the correspondence between mean level of confidence and mean level of performance. Relative accuracy (also known as resolution) refers to the ability to discriminate items that will be remembered correctly or incorrectly from one another on an item-by-item basis. In this paper, we review how accuracy is measured and then apply accuracy analyses to both judgments of learning and to tip-of-the-tongue states. We also consider the underlying mechanism that drives both judgments of learning and tip-of-the-tongue states, focusing on the research that demonstrates the important of inferential mechanisms for both. We also discuss the neural underpinnings of metamemory.
