Apply Spaced Repetition

---
name: apply-spaced-repetition
description: Use when designing a learning or study system that schedules review sessions to maximize long-term retention with minimum time investment
source: Ebbinghaus "Über das Gedächtnis" (1885) forgetting curve; Cepeda et al. "Distributed Practice in Verbal Recall Tasks" Psych Bulletin (2006) meta-analysis; Kornell & Bjork JEPLMC (2008)
tags: [spaced-repetition, memory, retention, learning-science, study]
verified: true
---

# Apply Spaced Repetition

Schedule review of material at expanding intervals timed to retrieve knowledge just before it would be forgotten, producing significantly higher long-term retention than massed practice.

## Why This Is Best Practice

**Adopted by:** Anki (40M+ users), Duolingo (600M+ users), medical and language education curricula worldwide, US military language training programs
**Impact:** Cepeda et al. (2006) meta-analysis of 317 experiments shows spaced practice produces 200–300% better long-term retention than massed practice for the same total study time; Ebbinghaus forgetting curve shows 70% of material is forgotten within 24h without review
**Why best:** The spacing effect is one of the most replicated findings in cognitive psychology — distributing review sessions produces durable memory traces that massed ("cramming") practice cannot achieve at equivalent time investment.

Sources: Ebbinghaus "Über das Gedächtnis" (1885); Cepeda et al. Psychological Bulletin (2006); Kornell & Bjork JEPLMC (2008); Pimsleur "Memory Schedule" (1967)

## Steps

1. **Identify target knowledge** — define the specific facts, concepts, vocabulary, or procedures to be retained long-term; spaced repetition is most powerful for declarative knowledge (facts, vocabulary, formulas).
2. **Break content into discrete items** — each flashcard or review item must be atomic — one concept, one fact, one definition per item; compound items confound retrieval and create artificial dependencies.
3. **Create the review queue** — enter items into a spaced repetition system (Anki, RemNote, SuperMemo) or build a manual schedule; initial review happens within 24h of first exposure.
4. **Apply the expanding interval schedule** — standard intervals: Day 1 → Day 2 → Day 4 → Day 8 → Day 16 → Day 35 → Day 90; each successful retrieval doubles the next interval; failed retrieval resets to Day 1.
5. **Rate retrieval difficulty honestly** — after each recall attempt, rate difficulty (Anki: Again/Hard/Good/Easy); dishonest "Easy" ratings accelerate the interval beyond retention, causing future failures.
6. **Review daily — not in bulk** — 15–20 minutes of daily review outperforms 3-hour weekend cramming for long-term retention; the system only works if review sessions happen at scheduled intervals.
7. **Prioritize new items deliberately** — limit new item introduction to 10–20 per day; adding too many new items creates review pile-ups 2–4 weeks later that overwhelm the daily session.
8. **Write effective retrieval cues** — bad card: "What is photosynthesis?" (too broad); good card: "What is the specific product produced by the light-dependent reactions of photosynthesis?" (atomic, retrieval-focused).
9. **Use cloze deletion for connected knowledge** — cloze (fill-in-the-blank) cards work better for relational knowledge than simple Q&A: "The mitochondria is the _____ of the cell" is more retrieval-effective than "What does the mitochondria do?"
10. **Review the review system** — after 3 months, assess item difficulty distribution; items stuck in "Again" repeatedly need to be rewritten as simpler atomic items; items never forgotten may be too easy to be worth reviewing.

## Rules

- Retrieval must be active — reading a card without attempting recall produces no spacing effect; always attempt to retrieve before revealing the answer.
- Consistency is more important than session length — 10 minutes daily is dramatically more effective than 70 minutes once per week.
- Items must be reviewed even when "you know it" — intuition about knowledge reliability is poorly calibrated; the system knows better than subjective confidence.
- New content must be encoded before spacing begins — spaced repetition reviews; it does not teach; initial encoding must happen through instruction, reading, or practice first.
- Do not import pre-made decks without reviewing them — generic decks often contain poorly written items; curate the items that are well-matched to retrieval need.

## Common Mistakes

- **Adding too many items per day** — adding 50+ new items creates an overwhelming review queue 3–4 weeks later; limit to 10–20 per day maximum.
- **Skipping review sessions** — a missed day multiplies future load; a missed week creates a near-impossible catch-up; treat daily review as fixed time.
- **Writing over-broad cards** — "Explain the French Revolution" cannot be retrieved in 10 seconds; atomic cards enable the system to function.
- **Using for procedural skills without practice** — spaced repetition maintains knowledge of how to do something; the doing still requires practice; they are complementary, not substitutable.
- **Front-loading a semester then stopping** — spaced repetition only retains what is actively reviewed; stopping review 2 months before an exam eliminates most of the benefit.

## When NOT to Use

- Skills requiring physical practice (driving, surgery, sports) — use deliberate practice instead
- Creative or generative knowledge where recall of fixed facts is not the primary performance requirement
- Very short-term knowledge requirements where forgetting after the event is acceptable