Spaced Repetition

Reference entry · last updated 20260910

Previous version, captured 20260910

Spaced repetition distributes repeated study or retrieval across separate sessions. Intervals may be fixed or adjusted according to recall performance and the intended retention period.[1, 2, 5]

First principles and definitions

The spacing effect is the retention advantage of distributing learning over time compared with concentrating it into closely grouped repetitions. Retrieval practice means attempting to recall information from memory. Spacing describes when practice occurs; retrieval describes the activity.[1, 2]

Illustrative example: rereading a definition on Monday and Thursday spaces study. Closing the text and recalling the definition on those days combines spacing with retrieval practice.

The study gap separates learning sessions. The retention interval separates the final learning session from a later test. A useful schedule depends partly on how long the material must be retained.[1]

Evidence and limits

In a study of facts tested up to one year later, Cepeda and colleagues found that the gap associated with the best final recall increased as the retention interval increased. Excessively short or long gaps reduced performance under the tested conditions.[1]

Increasing intervals are one scheduling choice. In Karpicke and Roediger's word-pair experiments, equally spaced retrieval produced better recall after two days than expanding retrieval. Expanding retrieval performed better on a short-delay test in their first experiment.[2]

Retrieval practice also depends on the test delay. Roediger and Karpicke found that recalling prose passages without feedback improved retention after two days or one week compared with repeated study. Repeated study produced better performance after five minutes.[3]

Explanations of the spacing effect

Proposed explanations include changes in the context encoded during repetitions and retrieval of an earlier learning episode during later study. Karpicke and Roediger discuss retrieval difficulty as an explanation for the benefits of spaced testing. Such accounts must be distinguished from the observed retention results; the experiments do not establish one mechanism for every spacing task.[1, 2]

Scheduling approaches

Leitner-style boxes: cards move between groups according to recall success. Cards recalled correctly advance; forgotten cards return to the first group. The original system used box position to prioritize practice. Fixed daily or weekly schedules are later adaptations.[4]
SM-2: successful repetitions start with intervals of one and six days. Later intervals multiply the previous interval by an item-specific ease factor, updated from a 0–5 response grade and bounded below at 1.3. Grades below 3 restart the repetition sequence while retaining the updated ease factor. The original procedure also repeats grades below 4 within the session.[5]
Anki's FSRS: the Free Spaced Repetition Scheduler uses review history to model memory and schedule reviews toward a chosen desired retention. Higher desired retention generally means shorter intervals and more reviews. The setting is a scheduling target, not a guarantee for every card or learner.[6]

Applications

Michael Nielsen's 2018 essay Augmenting Long-term Memory describes his use of Anki for reading and learning technical subjects. It offers practitioner experience and examples of question design.[7]

Andy Matuschak and Nielsen's Quantum computing for the very curious embeds spaced review questions in an explanation of quantum computing. This mnemonic medium combines exposition with recurring recall prompts. It is a design example; claims about its effects require separate evaluation.[8]

References

Nicholas J. Cepeda, Edward Vul, Doug Rohrer, John T. Wixted, and Harold Pashler. “Spacing effects in learning: A temporal ridgeline of optimal retention.” Psychological Science 19(11), 1095–1102, 2008. DOI: 10.1111/j.1467-9280.2008.02209.x.
Jeffrey D. Karpicke and Henry L. Roediger III. “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, 2007. DOI: 10.1037/0278-7393.33.4.704.
Henry L. Roediger III and Jeffrey D. Karpicke. “Test-enhanced learning: Taking memory tests improves long-term retention.” Psychological Science 17(3), 249–255, 2006. DOI: 10.1111/j.1467-9280.2006.01693.x.
Piotr Woźniak. “The true history of spaced repetition.” SuperMemo, June 2018. Vendor account of Leitner's 1972 box system and later adaptations.
Piotr Woźniak. “Algorithm SM-2.” Original SuperMemo algorithm specification.
Anki Manual. “FSRS” and “Desired Retention.”
Michael Nielsen. “Augmenting Long-term Memory.” 2018.
Andy Matuschak and Michael Nielsen. “Quantum computing for the very curious.” Quantum Country, 2019.