Strategic Studying: The Value of Forced Recall
Posted Oct 09, 2016
School has started, and many students are discovering that they are not doing as well as expected. Parents and teachers may be chiding them about working harder. It might be more helpful to urge them to work smarter. This brings us to the matter of how students study.
My impression is that many students do not study effectively. Everyone knows that it is a bad idea to try to study while listening to music, watching TV, or frequently interrupting to check e-mail or Facebook and Twitter. One aspect of studying that is often under-valued is the way students test themselves to see how much they have learned. Typically, they "look over" the assigned learning content (notes, on-line videos, or reading assignments). Most students do not realize how important it is to force themselves to recall. In part, this is because they are conditioned by multiple-choice tests to recall passively, that is recognize when a correct answer is presented, as opposed to generating the correct answer in the first place.
Studies of student learning practices reveal how important to memory formation it is to retrieve information you are trying to memorize. For example, a 2008 study evaluated study and testing effects on memorizing foreign-language word pairs in one learning session of four trials, as one might do for example with flash cards.
A large recall improvement occurred if each repeated study attempt required active recall at that time, as opposed to just looking at the correct definition. Applying this finding to all kinds of learning suggests that learners should force themselves to recall what they think they have learned. Just looking at content again and again may not promote long-term learning.
Next, the investigators wanted to know whether recall is affected by focusing only on the word pairs that were incorrectly recalled. This is equivalent in a flash-card scenario, to re-studying only the words that were missed in the previous attempt. The test groups involved Study (S)(looking at each word and its paired word) and Test (T)(forced recall of each word in the pair) for either all of 40 word pairs or just the word pairs that were not recalled in the previous trial. The learners ran through the deck four successive times.
At the end of this learning phase, students in each group were also asked to predict how many word pairs they would be able to remember a week later. It turned out that irrespective of the learning condition, predictions were inaccurate. This confirms my own experience that students are frequently poor judges of how much they know.
As for the effectiveness of initial learning, all four groups achieved perfect scores after four trials, with the largest improvement between the first and second trial. So that means they all learned the material. The issue at hand was how well they remembered when quizzed later. When given a test a week later, the two groups in which forced-recall testing was repeated in each study trial, final recall increased over the other two conditions by four standard deviations, ranging from 63 to 95% of correct recall a week later. Thus, it seems that forced-recall testing is more important for forming memories than is the studying. What this indicates is that learning occurs during forced-recall testing, and retrieval practice should be part of the initial study process.
In 2015, another group of researchers replicated these findings and further examined the effects of the varied spacing in the first study. That is, in the 2008 study, the two conditions where testing was repeated in each trial took more time because all 40 word pairs were tested. The second group of investigators was surprised that the earlier study seemed to diminish the importance of repeated studying, compared with repeated testing. One problem might have been that the original study design was "between subjects," where scores were averaged for students in different test conditions. This design meant that the elapsed time varied among the groups, because it took more time to complete four study cycles of all 40 word pairs and tested than it did when only non-recalled items were studied and/or tested. So this new study had a "within subjects" design in which every learner experienced all four ST conditions on 10 different word pairs.
The results replicated the earlier findings on the value of forced-recall testing. That is, the two groups that self-tested in each of the four study cycles had the most recall after one week. Moreover, the group that re-studied and re-tested all word pairs recalled about twice as many word pairs than did the group that only re-studied and re-tested non-recalled words. Thus it appears that restudying items that have been correctly recalled earlier is far from useless.
Both studies make it clear that how well a learner remembers soon after learning provides no assurance of how much will be remembered after a week (or longer) delay. In these studies, optimal learning occurred when an initial learning session included repeated study and forced-recall testing of all items at least four times in a row. Of course, we only have data for 40 items, and long-term memory might be affected differently for smaller or larger sets of learning material.
- Just looking over learning material can be ineffective for long-term memory.
- Right after learning an item of information, force yourself to recall it and check to see if you got it right.
- Conduct forced--recall testing of all information, not just the items that were previously recalled correctly.
Study should be strategic. These and other learning and memory aids are found in my inexpensive e-book on learning skills, Better Grades, Less Effort (Smashwords.com) or the more comprehensive book, Memory Power 101 (Skyhorse).
Karpicke, J. D., and Roediger, H. I. III (2008). The critical importance of retrieval for learning. Science. 319, 966-968.
Soderstrom, N. C. et al. (2016). The critical importance of retrieval—and spacing—for learning. Psychol. Science. Dec. 16. doi:10.1177/0956797615617778.