Spaced repetition is the practice of reviewing material at gradually increasing intervals over time, rather than repeating it over and over in a single sitting. It sounds like a small, almost obvious idea. It's also one of the most thoroughly tested findings in the entire history of psychology, confirmed by experiments running from 1885 to the present day, and it explains why a revision session that feels smooth and successful in the moment often produces surprisingly little that survives to exam day.

The 140-year-old discovery behind it

In 1885, German psychologist Hermann Ebbinghaus published Memory: A Contribution to Experimental Psychology, based on an extraordinary act of self-experimentation: he spent years memorising lists of meaningless syllables, BOK, YAT, KUF, and testing his own recall of them after different lengths of time. What he found became known as the forgetting curve: memory for new information drops away quickly at first, then more slowly, in a predictable, near-exponential pattern.

Ebbinghaus also noticed something more hopeful, a phenomenon he called savings. Even after material seemed completely forgotten, relearning it took measurably less time than learning it fresh the first time, evidence that some trace of the original memory survives even when it can no longer be consciously recalled. And critically, when he relearned the same lists after a gap, the forgetting curve on the second pass was flatter than the first: each act of reviewing faded material made the memory more durable than it had been before.

What makes this finding unusually solid, in a field where many famous results have not held up well under modern scrutiny, is that it keeps replicating. A 2015 study by Murre and Dros repeated Ebbinghaus's original experiment as closely as modern methods allow and found essentially the same forgetting pattern across retention intervals from twenty minutes to thirty-one days, 130 years after the original data was collected.

Why spacing works, not just that it does

The mechanism, established by later research building on Ebbinghaus's original work, comes down to a distinction between two properties of a memory: how easily it can be retrieved right now, and how durably it's actually stored for the future. Reviewing material while it's still fresh feels productive and boosts easy access to it briefly, but does comparatively little to strengthen the underlying memory. Waiting until a memory has partially faded, then successfully retrieving it anyway, requires more effort, and that effort is precisely what builds a stronger, more durable trace than an easy review would.

This is why spaced repetition and effortful retrieval are so tightly linked in the research: spacing works largely because it creates the conditions for a harder, more valuable act of remembering, not simply because time has passed between sessions.

How big is the effect, actually?

This isn't a marginal, hard-to-detect finding. Cepeda and colleagues' 2006 meta-analysis, pooling 254 studies and more than 14,000 participants, found distributed practice outperformed massed, concentrated practice for long-term retention across virtually every condition tested. A synthesis of meta-analyses by education researcher John Hattie found an average effect size of 0.71 for spacing, comfortably above the 0.4 threshold Hattie's own framework treats as a full year's worth of expected progress. And a study by Karpicke and Roediger found spaced retrieval practice produced recall roughly 150% better than the same material studied in a single massed session.

The debate even researchers haven't settled

It would be tidy to say the science recommends a specific schedule, review after one day, then three days, then a week, each gap a little longer than the last. The honest picture is more interesting than that. Landauer and Bjork's influential 1978 proposal argued that expanding intervals, each review spaced further apart than the last, should produce the strongest long-term retention. But Karpicke and Roediger's own research found the opposite for material retained over a longer period: expanding schedules helped short-term recall, but equally spaced review sessions produced better retention a week or more later, precisely the timescale that matters for a GCSE exam months away.

The resolution researchers have converged on is more specific and, usefully, simpler to apply than either camp's original claim: what matters most is that the very first review is delayed long enough to make retrieval genuinely effortful, not immediate or easy. Whether subsequent reviews expand, stay equal, or vary isn't nearly as consequential as making sure that first retrieval attempt happens after a real gap, not right after the material was first learned.

Spaced repetition in digital tools

The first practical system for applying this outside a research lab was remarkably low-tech: the Leitner system, developed by Sebastian Leitner in 1972, sorts flashcards into boxes reviewed at different frequencies, cards answered correctly move to a less-frequent box, cards answered incorrectly move back to a more-frequent one. Modern spaced repetition software, from Anki to language apps like Duolingo, automates the same underlying idea with algorithms, most descending from a system called SM-2, that estimate exactly when a specific fact is about to be forgotten for a specific learner and schedule the next review accordingly.

It's worth knowing what these tools are genuinely well-suited to, and where they run into limits. Flashcard-style spaced repetition excels at discrete facts, vocabulary, dates, formulae, definitions, where there's a clear right answer to retrieve. GCSE subjects usually demand more than that: understanding why a formula works, not just recalling it, or constructing an argument, not just remembering a fact. This is where a full adaptive learning platform differs from a pure flashcard app, applying the same spacing principle to genuine conceptual practice and worked problems, not only to isolated facts.

Building spaced repetition into a real revision timetable

Don't finish a topic and move on for good. The most common way spacing gets lost in ordinary revision is treating each topic as complete once it's been covered once. Build in a deliberate return to it a week or two later, even briefly, rather than assuming the first pass was enough.

Interleave topics rather than blocking them. A revision timetable that spends a whole week solely on one topic before moving to the next removes the natural spacing that comes from mixing subjects and topics across a week. Rotating between several topics within the same week creates spacing almost automatically.

Use a real calendar, not memory, to track what needs revisiting. Because the effect depends on returning to material after a genuine gap, some external structure, a written or app-based schedule marking when each topic was last covered, matters more here than in most other revision techniques, since the whole method depends on timing that's easy to lose track of otherwise.

Don't worry about the exact interval pattern. Given that the research itself hasn't settled whether expanding or equal intervals work better long-term, a reasonably spaced, roughly consistent schedule, revisiting each topic every one to three weeks through a term, is a perfectly reasonable approach, not a compromise on some more scientifically correct alternative.

This is precisely the mechanism built into BrainStrata's adaptive practice: every topic a student has covered gets automatically resurfaced after a calculated gap, calibrated to that specific student and topic, so the spacing this research has spent 140 years confirming happens by default, rather than depending on a revision timetable someone has to remember to build and maintain by hand.

Sources and further reading

[1] Ebbinghaus, H. (1885/1913). Memory: A Contribution to Experimental Psychology. The original documentation of the forgetting curve and the spacing effect.

[2] Murre, J. M. J., & Dros, J. (2015). Replication and Analysis of Ebbinghaus' Forgetting Curve. PLOS ONE, 10(7). Modern replication of Ebbinghaus's original findings across retention intervals of 20 minutes to 31 days.

[3] 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. Meta-analysis of 254 studies and over 14,000 participants.

[4] Hattie, J. (2009). Visible Learning: A Synthesis of Over 800 Meta-Analyses Relating to Achievement. Routledge. Reported mean effect size of d = 0.71 for the spacing effect.

[5] Karpicke, J. D., & Roediger, H. L. (2007). Expanding retrieval practice promotes short-term retention, but equally spaced retrieval enhances long-term retention. Journal of Experimental Psychology: Learning, Memory, and Cognition, 33(4), 704–719.

[6] Leitner, S. (1972). So lernt man lernen (How to Learn to Learn). Origin of the Leitner system, the first widely used practical spaced repetition method.