| Studying for an exam involves not only committing facts to memory, but evaluating what has been learned and how long it will be retained—metamemory processes. Previous research has shown that the accuracy of metamemory is improved as more information about the test questions is made available (see Maki & Serra, 1992). However, this has not been investigated using an expository text typical of an academic setting in a within-subjects design. In addition, the effect of time before the test (retention interval) on metamemory judgments has rarely been investigated. In four experiments, participants read a 4701 word psychology textbook chapter and predicted their test performance over as many as four retention intervals ranging from 10-minutes to 3-weeks. The timing of the judgments (before reading, after reading, and after the test) was also varied. Post-reading predictions were based on topics of the test questions, the actual questions, or no cues. Consistent with previous studies, test performance generally decreased in an Ebbinghausian function, with the majority of forgetting occurring within 48 hours. Performance on factual questions was better and more dependent upon knowledge acquired from reading than performance on applied questions. Judgments were higher when based on more information about the test questions (i.e., topic-based predictions and postdictions), but were more accurate only when made 3 weeks before the test. |
