Slowing the Process of Forgetting

Everyone has experienced the derailment of a train of thought or struggled with that tidbit of information on the tip of your tongue. Forgetting is a problem for many people and a nuisance to most of us. Wouldn’t it be wonderful to learn how to remember more?

Learning techniques that slow the process of forgetting were the focus of an invited address by APS Fellow Harold Pashler, University of California, San Diego, at the APS Annual Convention in Atlanta. In “Optimizing Resistance to Forgetting: What Does Psychological Science Have to Offer,” Pashler examined “how the degree and timing of instruction affects forgetting over periods ranging from days to weeks out to six months and more, using a variety of simple but realistic learning tasks.”

According to Pashler, the main focus of learning and memory research has been to bring about “momentary insight,” that teaches something until it has been learned. While this initial learning is important, he pointed out that with the passage of time people forget much of what they learned. Some schoolteachers, Pashler said, often describe the beginning of a new school year as “starting all over again,” re-teaching students everything they have forgotten over the summer.

Some problems with prior research in this area, according to Pashler, are that it typically uses highly irrelevant tasks, such as learning nonsense syllables, tests with little educational relevance (free recall, implicit tasks) and a very short retention interval, usually within the same day. In contrast, Pashler’s research team uses relevant, real-world learning tasks, such as language learning, with longer study-test intervals. Their goal is to determine the best study strategies to slow the process of forgetting and to make learning faster. They are attempting to discover the optimum study-study-test interval and whether feedback or overlearning methods are effective.

Pashler’s current studies have shown that testing knowledge, instead of just presenting it, works best. For example, after learning the Inuit word for walrus (kunivik), the best learning occurs when the next presentation is “walrus = __________” instead of “walrus = kunivik.” The research shows that complete feedback is essential to learning when the answer given is incorrect and that reinforcement of correct responses contributed little or nothing to learning.

The researchers also found that a longer test interval is best, even though it may produce errors at first. In the past, researchers thought that testing often was the best method, or that testing should be done just before the material is forgotten but before the participant is likely to make a mistake. But, Pashler and his colleagues have found that even though a longer test interval produces a greater number of errors, it is more effective at learning for retention.

Their final conclusion was that spaced practice is disadvantageous when the retention interval is short. For instance, cramming for a test the night before may not be as harmful as many would think because spaced study may actually be a worse strategy. The optimum study-study interval may be some ratio of the study-test, or retention, interval. According to Pashler, this number may be one or close to one, meaning that the time between study sessions should be approximately equal to the time between studying and testing, but research into exactly what that interval is has not yet been done.