How many times should you review something to remember it?
How learning works · Published · Evidence: Moderate · About this guide
How many times should you review something? One research group's suggestion is to recall each item correctly about three times in a first session, then relearn it three times at widely spaced intervals; it comes from studies of college students learning concepts [1]. It is a suggestion, not a guarantee: for problem solving the benefit was smaller [2], and how much practice very long-term memory needs has not been studied systematically [3].
Students very commonly stop studying something the moment they can answer it [4]. It feels finished. Research on practice suggests that the first correct answer is a useful checkpoint but a poor finish line, and that, for memory weeks or months later, practice spread over later days tends to pay off more than the same practice piled into one sitting.
Is getting it right once enough?
In a well-known experiment, university students learned 40 Swahili–English word pairs through rounds of studying and testing themselves [5]. Once a student had recalled a pair correctly, what happened next depended on the group: the pair stayed in both the study and the test rounds, was dropped from study but still tested, was dropped from testing but still studied, or was dropped from both. During learning the groups looked the same: by the end, nearly every student had recalled all 40 words at least once.
A week later they were far apart. Students who had kept being tested on pairs they already knew recalled about 80% of them. Students whose pairs were dropped from testing after the first correct recall recalled about a third (36% and 33%), even when they had kept re-reading those pairs. More studying after the first success added almost nothing; more recalling added a lot.
The students did not see this coming. In every group they predicted they would remember about half.
What is successive relearning?
Successive relearning is a name for a simple routine: practise recalling each item until you get it right, then come back in later sessions on other days and practise again until you get it right once more [6]. It joins two well-supported ideas: retrieval practice (what to do) and spacing (when to do it) [3]. In the successive-relearning studies below, each attempt was followed by the correct answer, so a wrong answer was corrected straight away instead of being left to stand.
How many correct recalls, and how many sessions?
A detailed answer comes from three experiments with 533 college students learning conceptual material [1]. Students practised each item until they had recalled it correctly between one and four times in a first session, then relearned it in one to five later sessions. Their memory was tested one to four months afterwards.
Two patterns stood out. Extra correct recalls in the first session helped at first, but that advantage shrank as relearning sessions were added. And relearning sessions gave long-term memory a large boost for relatively few extra practice trials. Weighing lasting memory against effort, the authors suggested that students recall each concept correctly three times in a first session, then relearn it three times at widely spaced intervals.
Relearning also gets quicker. Getting items right again in a later session usually takes much less time than learning them did the first time, so the extra sessions cost less than they sound [1][3].
What pays off less is piling more practice into one sitting, including overlearning: carrying on after you already get every item right. In two experiments, college students learned simple pairs (cities and their countries, or rare words and their meanings) through rounds of testing with feedback. Those who got much more of this practice per item in one sitting, and kept going after they were already getting every item right, remembered far more a week later than those who got much less, many of whom did not get every item right in any single round. The edge shrank quickly: it was no longer statistically significant at the four-week test in one experiment and at the nine-week test in the other [7]. The authors note that their comparison probably overstated, rather than understated, the benefit of the extra practice. Extra drilling in one sitting mostly buys a short-term gain.
Does it work outside the lab?
In a real introductory psychology course, students used a computer flashcard program to practise some of the course concepts this way, retrieving each one correctly in several sessions spread over days. On the course exams, they scored over 10% higher on those concepts than on concepts they studied in their own way, even though the exam asked them to apply the concepts rather than recite definitions [8][3]. On a recall test 24 days after the exam, students remembered over 60% of the relearned concepts and under 20% of the ones they had learned on their own. Another group, which spent the same spaced sessions re-reading instead of recalling, also gained, but less.
These comparisons did not control how long students spent with each concept. The authors of a later course study argue that comparisons like these may have considerably overestimated the effect [9].
That later study controlled how long students saw the material under each method, and the same students switched between the methods from week to week. On recall tests six to ten weeks after the lectures, students remembered more of the very items they had practised with successive relearning than of items they had restudied. On questions that asked for a different part of a practised statement, or about material from the same lectures that had not been practised, the gain was small and only borderline significant [9].
The method has limits. When college students learned to solve probability problems, reaching a correct solution in each of three sessions beat reaching the same number of correct solutions in one session, but only by a small margin, and scores a week later stayed at about half or less [2].
Can you trust your own sense of "I know this"?
Not on its own. When students were allowed to set aside flashcards they felt they knew, they learned slightly but consistently less than when every card stayed in play [4]. Dropping cards seems sensible because it frees time for the harder ones, but it rests on two shaky foundations: how well you judge what you know, and how much you think further practice of "known" items is worth.
Checking your own answers matters too. In two studies, college students practised definitions until they had judged their own answer correct three times. Those who scored their answers too generously stopped practising definitions they had not really learned, and the less accurately students judged, the less they remembered on a test two days later [10]. In the same research, comparing a written answer point by point with the key ideas of the correct answer helped students judge more accurately.
Is a pass mark the same as mastery?
"Mastery learning" means moving on once you reach a set level, with extra help on what you missed. A meta-analysis of 108 evaluations found that mastery programmes raised exam performance in college, high school and the upper elementary grades, by about half a standard deviation on average and more for weaker students [11]. The effects varied with how the programmes were run, and were much smaller on standardized tests than on tests written for the course. Mastery programmes could also take more time, and self-paced versions often lowered completion rates in college classes.
You may have heard a bigger number. Benjamin Bloom reported that students tutored one-to-one with mastery-style correction scored about two standard deviations above a regular class: the average tutored student outscored 98% of that class [12]. This "2 sigma" result came from the dissertations of two University of Chicago doctoral students. A later review of tutoring experiments put the effect of human tutoring at about 0.8 standard deviations [13]. Bloom's figure is a challenge worth chasing, not a typical result.
A pass mark is a sensible point to stop drilling for today. It does not show that the material will still be there next month.
When can you stop studying a topic?
Taken together, the research points to a rule of thumb rather than a formula:
- In a first session, keep going until you can produce each answer yourself. Getting it right more than once helps, though that gain shrinks once you come back to the items on later days.
- Check each answer against the correct one, not against a feeling of familiarity.
- Then stop for the day instead of drilling the same items over and over.
- Come back on another day and practise until each item is right again. Do this a few times.
- Expect later sessions to be shorter. Items you keep missing deserve the extra time.
What the research does not say
- "Three and three" is not a law. The suggestion comes mostly from one research group's studies with university students learning short concepts. For problem-solving the benefit was smaller [2].
- How much is enough for very long-term memory is open. The researchers themselves say this has not yet been studied systematically [3].
- It does not settle the spacing. These studies do not say how far apart your sessions should be, or show that steadily growing gaps beat equal ones. That is a separate question, covered in the article on spaced repetition.
- Most participants were adults. There is very little research on successive relearning with children [3].
- Averages are not predictions. Classroom gains are averages across many students; your own results will vary.
- A pass mark is not memory. Mastery programmes raised scores by about half a standard deviation on average [11], and one-to-one human tutoring by about 0.8 [13], not the two of Bloom's headline. Reaching a criterion once says little about next month without coming back to it.
Related articles
-
Retrieval practice (active recall)
Retrieval practice, or active recall: why recalling and then checking the answer tends to beat rereading, how large the testing effect is, and its limits.
-
Spaced repetition and spaced practice
The spacing effect: why spreading study over days helps memory last, how far apart reviews should be, and whether ever-longer intervals beat equal gaps.
-
Feedback and learning from mistakes
Why checking answers after a quiz matters, when a wrong guess helps you learn, and why confident mistakes are often easiest to fix, though they can return.
-
Exam study plan: planning your reviews before an exam
Start from the exam date: how far apart to space reviews, what to do in each one, and an example four-week revision plan, with the limits of the evidence.
-
The illusion of competence: when studying feels better than it works
Familiar is not the same as remembered. Why rereading and highlighting feel effective but rate low in research, and how to check what you really know.
-
Exam stress and memory
What stress does to recall in exams, what a well-known 2016 experiment found about practice tests under pressure, and how far later studies support it.
Sources
- [1] Rawson, K. A., & Dunlosky, J. (2011). Optimizing schedules of retrieval practice for durable and efficient learning: How much is enough? Journal of Experimental Psychology: General, 140(3), 283–302. experiment · doi.org/10.1037/a0023956 ↩ Back to the text
- [2] Rawson, K. A., Dunlosky, J., & Janes, J. L. (2020). All good things must come to an end: A potential boundary condition on the potency of successive relearning. Educational Psychology Review, 32(3), 851–871. experiment · doi.org/10.1007/s10648-020-09528-y ↩ Back to the text
- [3] Dunlosky, J., Greve, M., Badali, S., Wissman, K. T., & Rawson, K. A. (2023). Successive relearning: An introduction and guide for educators. In C. E. Overson, C. M. Hakala, L. L. Kordonowy, & V. A. Benassi (Eds.), In their own words: What scholars and teachers want you to know about why and how to apply the science of learning in your academic setting (pp. 83–93). Society for the Teaching of Psychology. book chapter · teachpsych.org ↩ Back to the text
- [4] Kornell, N., & Bjork, R. A. (2008). Optimising self-regulated study: The benefits—and costs—of dropping flashcards. Memory, 16(2), 125–136. experiment · doi.org/10.1080/09658210701763899 ↩ Back to the text
- [5] Karpicke, J. D., & Roediger, H. L., III (2008). The critical importance of retrieval for learning. Science, 319(5865), 966–968. experiment · doi.org/10.1126/science.1152408 ↩ Back to the text
- [6] Rawson, K. A., & Dunlosky, J. (2022). Successive relearning: An underexplored but potent technique for obtaining and maintaining knowledge. Current Directions in Psychological Science, 31(4), 362–368. review · doi.org/10.1177/09637214221100484 ↩ Back to the text
- [7] Rohrer, D., Taylor, K., Pashler, H., Wixted, J. T., & Cepeda, N. J. (2005). The effect of overlearning on long-term retention. Applied Cognitive Psychology, 19(3), 361–374. experiment · doi.org/10.1002/acp.1083 ↩ Back to the text
- [8] Rawson, K. A., Dunlosky, J., & Sciartelli, S. M. (2013). The power of successive relearning: Improving performance on course exams and long-term retention. Educational Psychology Review, 25(4), 523–548. experiment · doi.org/10.1007/s10648-013-9240-4 ↩ Back to the text
- [9] Higham, P. A., Zengel, B., Bartlett, L. K., & Hadwin, J. A. (2022). The benefits of successive relearning on multiple learning outcomes. Journal of Educational Psychology, 114(5), 928–944. experiment · doi.org/10.1037/edu0000693 ↩ Back to the text
- [10] Dunlosky, J., & Rawson, K. A. (2012). Overconfidence produces underachievement: Inaccurate self evaluations undermine students' learning and retention. Learning and Instruction, 22(4), 271–280. experiment · doi.org/10.1016/j.learninstruc.2011.08.003 ↩ Back to the text
- [11] Kulik, C.-L. C., Kulik, J. A., & Bangert-Drowns, R. L. (1990). Effectiveness of mastery learning programs: A meta-analysis. Review of Educational Research, 60(2), 265–299. meta-analysis · doi.org/10.3102/00346543060002265 ↩ Back to the text
- [12] Bloom, B. S. (1984). The 2 sigma problem: The search for methods of group instruction as effective as one-to-one tutoring. Educational Researcher, 13(6), 4–16. review · doi.org/10.3102/0013189X013006004 ↩ Back to the text
- [13] VanLehn, K. (2011). The relative effectiveness of human tutoring, intelligent tutoring systems, and other tutoring systems. Educational Psychologist, 46(4), 197–221. review · doi.org/10.1080/00461520.2011.611369 ↩ Back to the text