Transitions between short-term and long-term memory in learning meaningful unrelated paired associates using computer based drills

摘要

Two computer-based instructional design experiments, modeled after the traditional flash card drill, investigated optimum queuing for reintroducing missed items in paired associate learning. In the first investigation, the number of items intervening between a word and its later review (the insertion gap) was manipulated to determine the drop-off points (i.e., the points at which information fades from memory). Data from Experiment 1 showed that the use of two intervening items resulted in significantly higher correct recall (P <.001) than the use of 3, 4, 5, or 6 intervening items. Furthermore, contrast coefficients confirmed two intervening items as the drop-off-point. The second experiment took the findings from the first experiment as a constant, using two intervening items as the first insertion gap. The number of intervening items in the second insertion gap (second drop-off point) was manipulated. A long-term retention test was administered seven days later. With minimal cues, no group was more impervious to inaccessibility over time than any other group, and after a prompted reminder the groups reinstated their original levels of learning. Experiment 2 showed that the use of 2 or 3 intervening items resulted in significantly higher correct recall than the use of 4, 5, or 6 items (three intervening items were slightly better than two). Although prior research had indicated that the optimal insertion gaps for a missed item are 1, 3, and 5 later insertions, the present experiment demonstrates that a treatment that uses fewer, specifically two or three intervening items is more beneficial than one which arbitrarily uses varied schedules for review.