Skip to content

Spaced Learning

Pattern · All patterns
Evidence · 54 claims (25 for, 26 mixed, 3 against) · 39 studies (18 causal, 8 review, 7 quant-synthesis, 3 theoretical, 2 design, 1 qualitative), q1–q4 · 12 of 39 report an effect size · 49 claims rest on one study

Description

A reusable policy for distributing learning opportunities, preserving total practice, gap and assessment horizon when comparing schedules.

Principle

This reusable policy instantiates the spaced learning principle.

Learner state, objective and valued goal

This is a reusable design model. A study result supplies a bounded observation; it does not reveal the state or future outcome of a new learner.

Before selecting a configuration, record the target task and representation, task-specific prior experience, access and language constraints, time available, observed responses and the exact help provided. Preserve an unknown rather than replacing it with a generic label such as novice. Ask the learner what they want to accomplish and why it matters. Record that valued goal separately from the designer's learning objective; agreement is an observation to elicit, not an assumption. If they diverge, negotiate the task or purpose and retain the unresolved difference.

Define the intended capability, a representative assessment task and scoring instrument, the criterion chosen for this design, and the assessment horizon. Agree what response would warrant changing support and how to check later independent performance. Criteria are local decisions, not universal mastery thresholds. If the brief only asks for better learning, elicit these fields before selecting an exact dose or forecasting a gain.

Keep the record in this form: response under stated conditions → uncertain state hypotheses → discriminating observation → next activity → response at the stated horizon. Difficulty, accuracy, speed, confidence and engagement are different observations. None alone proves learning. Record a learner's changed purpose or constraints as well as a changed performance.

Observation and adaptation

This diagnostic policy is a design proposal, not a validated scheduling algorithm. Start with an unassisted sample of the target items, record item-level errors and the time since last exposure, then separate cued recognition, recall and meaningful use. Record restudy, corrective feedback, retrieval opportunities and actual gaps; changing them together changes the intervention.

Observation Rival interpretations and discriminating observation Proposed next action
An item is not recalled Weak initial encoding and loss of access across the gap are different possibilities. Check comprehension/recognition with a cue and whether the learner can explain the association; retest unassisted after relearning and an interval. If the association is not understood, clarify and re-encode it. If understood but inaccessible, use corrective feedback and another separated retrieval. Do not optimize the gap from a single failure.
Immediate recall is fluent Recent repetition and durable retrieval both predict this. Use an unassisted check after the target delay. Maintain the delayed check rather than count fluency as retained learning; revise the schedule only after collecting later evidence.
An item is recalled correctly at its first spaced test A durable memory and a recent, easy retrieval both predict this; which one it is depends mostly on how long the gap was. In one experiment with undergraduates learning foreign-word pairs, dropping an item from practice once it was recalled reached about the same final recall in fewer trials when about 23 other items came between its presentations, but gave lower final recall when only 5 did (Long-gap dropout matched conventional schedules in fewer trials [+W]; Short-gap dropout lowered final recall [-W]; Dropout depends on the interval [~W]). Dropped items lost a little against keeping them at long gaps, offset by the extra practice missed items got (Dropping correct items cost a little, offset by gains on missed ones [~W]); in both experiments items missed at first were more often recalled at the final test under long-gap dropout than under either conventional schedule, which the authors attribute to the extra trials those items received, not to the gap or item difficulty (Dropout helped initially missed items most [+W]). Massed repeated recall after a first correct recall added nothing on a one-week test (Spaced retrieval wins at the final test despite lower initial recall [+M]). "About the same" there is a nonsignificant difference, not shown equivalence. If the gap before that recall was long, the item may leave this session's practice and the time go to missed items; keep it in the later spaced reviews. If the gap was short, keep practising it after a longer gap rather than dropping it. Do not repeat it immediately; massed repetitions add nothing.
The learner wants to skip reviews because the items feel known Feeling of knowing after recent exposure and durable recall predict the same report; learners judge massed study more effective even when their own delayed test shows spacing worked better (Learners misjudge spacing benefits [~M]). Asking learners to rate how likely they are to recall each item did not change the spacing benefit in one free-recall experiment, so collecting such ratings is not itself a cost (Likelihood ratings left the spacing effect unchanged [+W]). Compare the report with an unassisted check after a gap; the observation table on the principle gives the item-rating version of this row. Keep the scheduled reviews; let the delayed check, not the feeling, decide whether an item leaves the schedule.
Words are recalled but not used appropriately Item retrieval and contextual language use are different capabilities; ask for meaning and use in a new conversational context. Add contextual practice if recall succeeds but use fails. Do not claim that a word-pair recall study validates conversation transfer.

If six of ten words are recalled today, that is a sample response, not a calibrated probability of recall next week. For a learner valuing conversation, record the designer's recall criterion (for example eight of ten is a local choice), the sampled prompts and the one-week horizon; also agree how appropriate use will be judged in a representative conversation. Do not silently substitute recall for that valued goal. If these fields are missing, elicit them before choosing a calendar schedule.

Evidence-bounded expectation

A word-pair experiment separated total retrieval spacing from its relative arrangement: after first correct recall, repeated retrievals with intervening trials improved one-week recall relative to massed retrieval. At matched total nominal spacing, expanding, equal and contracting arrangements did not differ detectably: absolute versus relative spacing observation [~W]. The full source was checked. This does not establish equivalence or an advantage for expanding gaps, a universal calendar gap, or conversational use. Retain total practice, corrective feedback, initial learning criterion, actual gaps and final-test horizon when comparing schedules. Relearning to a criterion in later sessions breaks that rule: a longer gap means more forgetting and so more relearning trials, which confounds spacing with the amount of practice, so a local comparison should give a fixed number of review trials instead (Relearning-to-criterion designs confound spacing with practice amount [~W]; a methodological argument in a synthesis, not a tested result).

An equal schedule may be selected for practical reasons while this exact comparison remains unresolved; do not predict a gain from changing only equal to expanding gaps on this evidence. Calendar timing and the dose needed for a particular learner remain local questions. Lower practice fluency can coexist with better delayed performance, but difficulty alone does not identify the cause or predict the outcome.

Default schedule, labelled as a design proposal. No study in the wiki tested this combination as a whole; each part names its basis, and the principle carries the evidence for the relationship itself. Use it where the brief gives no reason to do otherwise, and replace any part the brief settles.

  1. Session 1. Introduce a capped set of new items (see the flashcard row below for why a cap), with one study pass and one recall attempt with corrective feedback per item. Separate an item's repetitions with other items rather than repeating it at once.
  2. Later sessions. Hold the total practice time a massed plan would have used, and spread it over at least two later sessions on different days (Spaced practice improves retention [+S]). Each review is a recall attempt with feedback; where feedback cannot be immediate, releasing scores and worked solutions later the same day is enough to keep the review (the classroom-course row below). Use equal gaps unless items will be retrieved many times (the expanding-intervals row below).
  3. Gap length from the horizon. In one large experiment with adults learning facts in two sessions, the best gap was about 20–40% of a one-week test delay and about 5–10% of a one-year delay, so roughly one to two days for a test a week away and three to five weeks for one a year away (Spaced practice improves retention [+S]). Treat these as a starting point to check, not a rule.
  4. Progression. Early sessions carry more new items and more feedback; later sessions are mostly reviews, with missed items getting more of the time than items already recalled after long gaps (the first-spaced-test row in the table above). Add a review close to the date of use.
  5. When to move on. An item is retired from the schedule only when it is recalled unassisted after the longest gap planned for it; the gap and the criterion are local choices recorded with the design.

Fitting the design to a situation

Ask for three facts when the brief omits them: when the learners will need the material (the delay to the test or use), how many sessions and minutes the schedule can hold, and whether reviews can be recall attempts with feedback.

If the brief says… Then change… Basis
A course of several weeks with a final exam Plan each topic's first study and at least one later review in a different week; keep the total practice time the same rather than adding sessions Spaced practice improves retention [+S]
The exam or use is months away Set longer gaps between first study and review for material needed later; review again closer to the date. Do not fix one gap for every course. In one experiment the best gap grew with the delay but shrank as a share of it (about 20–40% of a week, 5–10% of a year), and past the best gap recall declined slowly. The gap that suits a distant test can hurt a near one: in a lab experiment with word lists, repeating words a week apart rather than on the same day lowered recall on an immediate test (.24 against .42) but raised it on a test a week later (.14 against .06), so a course with both an early quiz and a late exam needs both short and long gaps Spaced practice improves retention [+S]; The optimal gap grows with the retention interval [+M]; The gap's effect depends on the retention interval [~M]; Short spacing won at a short test, long spacing at a long one [~M]
Reviews can be short quizzes in an app or LMS Make each review a recall attempt with corrective feedback rather than a reread, and space the quizzes rather than giving them together. The meta-analysis behind the first two links (29 studies, g = 0.74) compared spaced with massed retrieval, so it shows spacing matters within retrieval; recall over rereading rests on the testing-effect evidence Spaced retrieval improves retention [+M]; Spacing out retrieval practice beats massing it [+M]; Retrieval practice improves long-term retention [+S]
The brief asks for a choice between expanding and fixed intervals Default to equal intervals for simplicity; consider expanding only if each item will be retrieved many times. Across 54 effect sizes the two did not differ (g = 0.034); studies giving more retrievals per item tended to favour expanding, a between-study moderator with no printed size, not a within-study test. An earlier synthesis of mostly paired-associate studies found expanding gaps ahead overall but called the result tentative because the studies varied widely, and a word-pair experiment found a 1-5-9 expanding schedule no better than 5-5-5 Expanding and uniform schedules do not differ [~M]; More exposures favour the expanding schedule [~M]; Expanding gaps led in an earlier synthesis, tentatively [~W]; Expanding was no better than fixed in one experiment [~W]
A flashcard or Leitner-box system with a steady flow of new items Cap the number of new items introduced per session; above a threshold, items pile up in the first deck and learning stalls. Give the lower decks (new and recently missed items) more review time than the higher ones. The threshold comes from a queueing model and one crowd-worker vocabulary experiment; the deck allocation is model output only, so tune the rate locally by watching how many items pile up in the first deck Too high a new-item rate causes a phase transition in learning rate [~M]; The optimal schedule reviews lower decks more often [+W]
Session time is short and the item list is long Drop an item from the session's practice once it is recalled, but only where other items separate its repetitions by a long gap; spend the saved trials on missed items. With short gaps, keep the item in practice. The saved trials pay off mainly on the items missed at first. One lab study of undergraduates with foreign-word pairs; untested in classrooms or apps in the wiki Long-gap dropout matched conventional schedules in fewer trials [+W]; Dropout helped initially missed items most [+W]; Dropout depends on the interval [~W]; Short-gap dropout lowered final recall [-W]
A one-day workshop with no follow-up sessions Add spaced follow-up quizzes in the weeks after the workshop Spaced retrieval after a financial workshop improved retention at five months [+M]
A taught course that keeps moving to new topics (mathematics, a cumulative curriculum) with no spare class time Give a short weekly recall quiz on a fixed set of earlier fundamental concepts, cycling so each returns about every four weeks, in place of part of the quizzing on current material. If feedback cannot be immediate, release scores and worked solutions later the same day. In one randomised trial in 17 precalculus sections (128 students), this raised a retention index on those concepts six months later (81.6% against 77.3%) and the whole post-test (72.4% against 68.5%), while the final exam on current material did not differ detectably (74.9% against 76.4%; equivalence not tested; no effect size printed). The benefit did not vary detectably with instructor experience or class hour, which is a non-significant test, not shown sameness Weekly spaced recall quizzes improved six-month retention [+M]; Spaced recall raised the post-test without harming new learning [+M]; Spaced recall quizzes helped without immediate feedback [+W]; The benefit held across instructor experience [+W]; The benefit held across class hours [+W]
online-self-paced with learners choosing when to study Release review sessions on a schedule and send reminders; do not leave the timing to learners, who prefer massing and, by a review's account, underuse spacing generally Learners misjudge spacing benefits [~M]; Learners underuse spacing and retrieval practice [+W]
Learners want saved problems back later rather than straight after solving Offer a later, collected review of saved or missed problems (a weekly "wrong-book") rather than an immediate follow-up card Learners rarely opened immediate transfer cards and asked for delayed review [~W]
Second-language vocabulary in a course of a semester or longer Space vocabulary reviews across the term; do not stretch the gaps across years Distributed practice benefits L2 learning, with limits on very long spreads [~W]
The schedule is judged by practice accuracy Report a delayed test as well; spaced reviews lower practice accuracy while raising delayed recall Spaced retrieval wins at the final test despite lower initial recall [+M]
The skill is a complex task or a motor skill rather than recall of items Still space practice, but expect a smaller benefit than for verbal material, and do not lengthen gaps as far as for facts: in a meta-analysis across task types the effect (d = 0.46 overall) was smaller for complex tasks, and longer gaps helped verbal tasks but reduced the benefit for skills such as typing Spaced practice improves retention [~M]
The learners are preschool children Rely on repetition itself and check recall at a delay before claiming a spacing gain; the preschool row under Fitting the design to a situation on the principle gives the evidence and its limits (lag within one list, no day-scale test at this age in the wiki) Preschoolers showed no spacing effect in free recall [~M]; Grade and spacing interacted in children's recall [~M]

The evidence by question

Further evidence, not yet read against this model

Claims this page cited before it was rewritten as a conditional model, or that were attached later. All thirteen listed here on 2026-10-19 (twelve distinct claims; the spaced-practice claim was listed twice under two old titles) were read against their Evidence entries that day, and every one bears on the model, so each now sits in it: the spaced-practice claim in the default schedule, the months-away row and the new complex-or-motor-skill row; the two Latimier et al. (2021) spaced-against-massed retrieval claims in the quiz row, which now says that meta-analysis compared spaced with massed retrieval, not retrieval with rereading; the two Latimier schedule claims in the expanding-intervals row, with the exposures moderator described as a between-study moderator; Karpicke (2017) in the practice-accuracy row and the first-spaced-test observation row; the two Leitner claims in the flashcard row and the evidence by question; the three Pyc and Rawson (2007) dropout claims in the new short-session row, the first-spaced-test observation row and the evidence by question; and the misjudgement claim in the self-paced row, a new observation row and the evidence by question. None is left here.

Twenty-one more were linked to this page by the orphan-claim linker after that and were read against their Evidence entries the same day (wave 2, 2026-10-19). Eleven now sit in the model: the Lindquist et al. (2024) precalculus trial's four claims in the new taught-course row (beside the trial's main claim, linked there too) and the default schedule's feedback note; the two older expanding-against-fixed claims in the expanding-intervals row; Glenberg and Lehmann's (1980) proportionality result in the months-away row; the Pyc and Rawson (2007) initially-missed-items claim in the short-session row and the first-spaced-test observation row; Toppino and DiGeorge's (1984) grade-by-spacing interaction in the preschool row; Zechmeister and Shaughnessy's (1980) rating result in the skip-reviews observation row; the relearning-to-criterion confound under the evidence-bounded expectation; and the underuse review in the self-paced row. The spaced-practice claim was already in the model. The nine below were read and do not change the schedule.

Elements

Source verification and open tests

Karpicke & Bauernschmidt (2011), author-hosted full text. Method and results checked for the bounded contrast above.

The examples are local design instances, not additional observations from those studies. Test the proposed interpretation and activity branches with new learner responses; compare competing designs under matched conditions before claiming a causal or predictive advantage. Reassess the model when the observations disagree with it.

Claims

Examples

Key Sources

  • Karpicke, J. D., & Bauernschmidt, A. (2011). Spaced retrieval: Absolute spacing enhances learning regardless of relative spacing. Journal of Experimental Psychology: Learning, Memory, and Cognition, 37(5), 1250–1257. doi:10.1037/a0023436
  • Cepeda, N. J., Pashler, H., Vul, E., Wixted, J. T., & Rohrer, D. (2006). Distributed practice in verbal recall tasks: A review and quantitative synthesis. Psychological Bulletin, 132(3), 354–380. doi:10.1037/0033-2909.132.3.354
  • Latimier, A., Peyre, H., & Ramus, F. (2021). A Meta-Analytic Review of the Benefit of Spacing out Retrieval Practice Episodes on Retention. Educational Psychology Review, 33(3), 959–987. doi:10.1007/s10648-020-09572-8
  • Karpicke, J. D. (2017). Retrieval-Based Learning: A Decade of Progress. Learning and Memory: A Comprehensive Reference, 2nd edition, Volume 2. https://doi.org/10.1016/B978-0-12-809324-5.21055-9
  • Kornell, N. (2009). Optimising learning using flashcards: Spacing is more effective than cramming. Applied Cognitive Psychology, 23(9), 1297–1317. doi:10.1002/acp.1537
  • Reddy, S., Labutov, I., Banerjee, S., & Joachims, T. (2016). Unbounded Human Learning: Optimal Scheduling for Spaced Repetition. KDD ’16, San Francisco, CA, USA. https://doi.org/10.1145/2939672.2939850
  • Cepeda, N. J., Vul, E., Rohrer, D., Wixted, J. T., & Pashler, H. (2008). Spacing effects in learning: A temporal ridgeline of optimal retention. Psychological Science, 19(11), 1095–1102. doi:10.1111/j.1467-9280.2008.02209.x
  • Donovan, J. J., & Radosevich, D. J. (1999). A meta-analytic review of the distribution of practice effect: Now you see it, now you don't. Journal of Applied Psychology, 84(5), 795–805. doi:10.1037/0021-9010.84.5.795
  • Kornell, N., & Bjork, R. A. (2008). Learning Concepts and Categories: Is Spacing the "Enemy of Induction"? Psychological Science, 19(6), 585–592. doi:10.1111/j.1467-9280.2008.02127.x
  • Pyc, M. A., & Rawson, K. A. (2007). Examining the efficiency of schedules of distributed retrieval practice. Memory & Cognition, 35(8), 1917–1927. https://link.springer.com/article/10.3758/BF03193420
  • Lindquist, D. S., Sparrow, B. E., & Lindquist, J. M. (2024). Spaced recall reduces forgetting of fundamental mathematical concepts in a post high school precalculus course. Instructional Science, 52, 859–878. doi:10.1007/s11251-024-09680-w
  • Glenberg, A. M., & Lehmann, T. S. (1980). Spacing repetitions over 1 week. Memory & Cognition, 8(6), 528–538. doi:10.3758/bf03213772
  • Toppino, T. C., & DiGeorge, W. (1984). The spacing effect in free recall emerges with development. Memory & Cognition, 12(2), 118–122.
  • Zechmeister, E. B., & Shaughnessy, J. J. (1980). When you know that you know and when you think that you know but you don't. Bulletin of the Psychonomic Society, 15(1), 41–44. doi:10.3758/bf03329756
  • Carpenter, S. K., Pan, S. C., & Butler, A. C. (2022). The science of effective learning with spacing and retrieval practice. Nature Reviews Psychology. doi:10.1038/s44159-022-00089-1
  • Nakata, T. (2015). Effects of expanding and equal spacing on second language vocabulary learning: Does gradually increasing spacing increase vocabulary learning? Studies in Second Language Acquisition, 37(4), 677–711. doi:10.1017/s0272263114000825
Page metadata
Field Value
type pattern
id spaced-learning
title Spaced Learning
description A reusable policy for distributing learning opportunities, preserving total practice, gap and assessment horizon when comparing schedules.
status review
grain_size course
canonical True
generated by claude/unspecified on 2026-10-19
trust unverified — no one has confirmed this page's content

Frontmatter is the page's machine-readable record — see the Schema & Guide.