Spaced Repetition
also known as spaced practice, distributed practice, Leitner system, SRS
Spaced repetition exploits the forgetting curve: material reviewed just before it would be forgotten is remembered more durably than material reviewed when it is still fresh. Review sessions are scheduled at expanding intervals — short initially, then progressively longer as each item becomes well-retained. This achieves long-term retention with far less total study time than massed practice. The Leitner box system is the classic low-tech implementation; modern software (Anki, SuperMemo) automates the scheduling algorithm.
How the learner advances
Intent. Maximise long-term retention of a large item set by scheduling each review at the latest moment before forgetting, systematically expanding the interval as retention strengthens.
When to apply. Apply whenever the goal is durable long-term recall of a large discrete item set — vocabulary, clinical drug interactions, protocol steps, API signatures, or any declarative knowledge base. Particularly valuable when the learner has limited daily study time and must retain material for months or years rather than for an exam in the next week. Not well suited to procedural skills or conceptual understanding that requires application rather than recall.
Threshold — earns the next step. The learner can retrieve any item in the deck with high accuracy on a cold test administered at least one week after the last scheduled review session for that item.
Masterpiece — the artifact that proves it. A mature spaced repetition deck — demonstrable by showing that the items in it are scheduled weeks to months out, and that a sample cold test achieves above 90% recall — evidence that the intervals have been earned, not assumed.
Facets
- Container — async
- Mode — self-pacedflashcardsolo
- Reach — individual
- Persona — learner
- Craft (AI Fluency) — diligence
- Learner — human
Inputs
- Item set with question-answer pairs — A collection of discrete facts, definitions, or associations structured as question-and-answer cards. Each item must be small enough to be retrieved in a single retrieval act.
- Scheduling mechanism — Either a physical Leitner box set with labelled intervals, or spaced repetition software (Anki, SuperMemo, Duolingo) that implements an SM-2 or similar algorithm to compute the next review date for each item.
- Daily practice commitment — A consistent short daily session (10–30 minutes) for reviews. Spaced repetition fails when sessions are missed because the items due pile up and the spacing calculus breaks down.
Outputs
- More capable learner — A learner with durable, retrievable knowledge of the full item set — able to recall items months or years after initial study with minimal refresher effort.
- Optimised review queue — A dynamically scheduled queue that surfaces exactly the items that are at risk of being forgotten, neither wasting time on well-retained items nor letting fragile items decay past recall.
Steps (5)
Create or import an item set
Break the knowledge domain into discrete question-answer pairs. Keep each card atomic — one question, one unambiguous answer. Complex concepts should be decomposed into multiple cards rather than crammed into one.
producesitem deck
Introduce new items in small batches
Add new items to the active deck gradually — typically 10–20 per day — rather than all at once. Flooding the deck creates an unmanageable initial review burden.
producesactive deck
Review daily at the scheduled interval
Each review session presents items due today based on the scheduling algorithm. For each item, attempt recall before revealing the answer. Judge your recall honestly — incorrect or uncertain responses reset the interval; correct responses extend it.
producesupdated retention estimates
Grade and reschedule
After each response, grade it (again, hard, good, easy in Anki's model) and let the algorithm compute the next review date. Items answered correctly multiple times graduate to longer and longer intervals. Items frequently missed are reviewed more often.
producesscheduled next review per item
Maintain the habit until the item set is mature
Continue daily reviews until all items in the deck have reached long intervals (weeks to months). At that point the maintenance load drops dramatically. A well-matured Anki deck of 2,000 cards requires only 10–15 minutes of review per day.
producesmature item set with long intervals
Principles
- Review just before forgetting, not just after learning: the spacing effect means later review produces stronger retention than immediate review.
- Retrieval is the mechanism, not re-reading: the act of pulling the answer from memory — even with effort — strengthens the trace more than passively reviewing the material.
- Consistency beats intensity: a daily 15-minute session outperforms a weekly 2-hour session for long-term retention because the spaced intervals compound.
Known uses (3)
Anki — open-source SRS software — Anki (open source)
medical education, language learning Anki is the dominant open-source SRS platform; medical students in particular use pre-built shared decks covering pharmacology and anatomy.
Duolingo spaced repetition engine — Duolingo
language learning Duolingo's 2016 ACL paper describes a trainable spaced repetition model calibrated on half-life regression from user response data.
Sebastian Leitner — So lernt man Lernen (1972) — Herder Verlag
general education The original low-tech implementation using a physical five-box flashcard system.
Known failure modes (3)
- [anti-pattern:card-bloat]
Adding complex multi-part cards instead of atomic ones makes retrieval ambiguous and forces the algorithm to treat partial understanding as full recall.
- [anti-pattern:missed-sessions]
Skipping days lets overdue items pile up into an unmanageable backlog; learners then avoid reviewing because the queue is overwhelming, creating a vicious cycle.
- [anti-pattern:passive-re-read]
Revealing the answer before attempting retrieval defeats the testing effect that gives spaced repetition its power; recognition is not recall.
Related trainings (3)
- Retrieval Practice★★
Strengthen long-term memory traces by repeatedly retrieving material from memory rather than restudying it, exploiting the testing effect.
- Mastery Learning★★
Ensure most learners reach a high standard on each prerequisite unit before advancing, by treating time-to-mastery as the variable rather than the performance ceiling.
- Microlearning★★
Deliver one specific, measurable learning outcome in the shortest engagement sufficient to achieve it, accessible in the flow of work.
Sources (3)
“Ebbinghaus published his hypothesis in 1885 after studying the memorisation of nonsense syllables; the forgetting curve shows that without reinforcement approximately 50% of information is forgotten within an hour”
“In Leitner's original method, published in his book So lernt man Lernen, the schedule of repetition was governed by the size of the partitions in the learning box.”
“spaced repetition is an evidence-based learning technique that uses expanding retrieval intervals to improve long-term retention”
Provenance
- Added to catalog:
- Last updated:
- Verification status: verified