Spaced Repetition: The Memory Science Behind Lasting Vocabulary

💡 TL;DR: Without review, you forget roughly 70% of new information within 24 hours. Spaced repetition solves this by scheduling each review at the moment just before you forget, multiplying long-term retention by up to three times compared to cramming, according to research spanning 317 experiments.

Why Does New Vocabulary Disappear So Quickly?

You study 30 new Vietnamese words on Monday. By Thursday, more than two-thirds are gone. This is not a personal flaw: it is the normal operation of human memory. In the 1880s, German psychologist Hermann Ebbinghaus memorized thousands of nonsense syllables and tested his own recall at different intervals. What he plotted became one of the most cited graphs in all of psychology: the forgetting curve.

The curve is steep. Without a review, roughly 70% of new material fades within 24 hours. Most of what survives that first day continues to decay over the next week until only a thin trace remains. The catch, which Ebbinghaus also discovered, is that memory is not permanent: it is a skill you maintain. Every time you successfully recall something, the curve flattens. The trace becomes stronger and persists longer before the next review is needed.

For language learners, this dynamic is both frustrating and useful. Frustrating because passive exposure, reading a word once or hearing it in a song, rarely survives the first night. Useful because there is a systematic fix.

What Is Spaced Repetition and How Does It Work?

Spaced repetition is the practice of scheduling reviews at increasing intervals, timed so that each review arrives just as the memory is about to fade. Instead of reviewing the same list every day (massed practice) or never reviewing at all, you revisit each item at the precise point where a bit of struggle is actually good for consolidation.

The mechanism works in two stages. First, retrieval practice: actively pulling a word from memory (rather than just re-reading it) strengthens the neural pathway more than passive study. Second, timing: reviewing at the edge of forgetting causes the brain to work slightly harder, which produces a stronger, longer-lasting memory trace. Researchers call this the desirable difficulty effect.

Polish computer scientist Piotr Wozniak turned this principle into an algorithm in the 1980s. His SM-2 formula, which still underlies the popular flashcard app Anki, calculates an "easiness factor" for each card and adjusts the next review interval based on how accurately you recalled it. A word you struggle with gets a short interval; a word you know cold gets pushed weeks or months into the future. The system quietly manages thousands of cards simultaneously.

What Does the Research Actually Show?

The spacing effect is one of the best-documented phenomena in memory science. A landmark 2006 review by Nicholas Cepeda and colleagues in Psychological Bulletin examined 839 effect-size contrasts from 317 experiments and found distributed practice reliably outperformed massed practice across every subject, age group, and format tested.

For second-language learners specifically, a 2022 meta-analysis by Kim and colleagues in Language Learning pooled 48 experiments involving 3,411 learners. Spaced practice showed a medium-to-large effect on vocabulary retention. Importantly, while shorter and longer spacing intervals performed similarly on immediate tests, longer spacing produced significantly better results on delayed tests, the ones that reflect real-world memory.

Perhaps the most striking demonstration of long-term spacing comes from a 1993 study by Bahrick and colleagues. They tracked foreign-language vocabulary over nine years and found that sessions spaced 56 days apart produced better retention than sessions spaced 14 or 28 days apart, even though the 56-day group had fewer total reviews. Longer gaps meant each review arrived when the memory truly needed reinforcement.

What Is the Optimal Gap Between Reviews?

Cepeda's 2008 follow-up study, which tested over 1,350 participants with retention intervals ranging from one week to one year, produced a practical rule: the optimal gap between reviews is roughly 10-20% of the time you want to remember something.

In practice, this means:

• Want to remember for 1 week: space reviews 1-2 days apart.
• Want to remember for 1 month: space reviews 3-6 days apart.
• Want to remember for 1 year: review every 2-7 weeks.
• Want to remember permanently: review every few months, with the gap expanding indefinitely.

Strategy	How it works	Short-term (1 week)	Long-term (6 months)
Massed practice (cramming)	Review everything in one block	High recall right after	Very poor retention
Fixed-interval spacing	Same gap every time (e.g., daily)	Good	Moderate; some overlearning, some forgetting
SRS (expanding retrieval)	Intervals grow based on recall quality	Slightly lower than cramming right after	Best: up to 3x vs massed
Contextual reading only	Encounter words naturally, no deliberate review	Highly variable	Poor for low-frequency words

SRS Apps: What Are Your Options?

You do not need to build your own scheduling system. Several tools implement spaced repetition well:

• Anki (free, open-source): the most customizable and research-faithful SRS app, used by medical students, language learners, and researchers. Steep learning curve, but highly effective.
• Duolingo: embeds a simplified spacing model, easy to start, but gives you less control over what gets reviewed and when. Good for beginners who need structure.
• Clozemaster: sentence-level flashcards for intermediate and advanced learners. Teaches words in context, which supports more natural recall.
• Migaku: integrates with video content, auto-generates cards from subtitles. Useful for immersion-based learning.

The app matters less than the habit. Any of these, used consistently, will produce better results than passive review or dictionary reading.

Cards, Sentences, or Reading: Which Combination Works Best?

Basic flashcards (word: translation) build recognition speed and are efficient for bootstrapping a vocabulary base. Sentence cards, where you see a full sentence with one word highlighted, train words in syntactic context, which is closer to how you will actually use them. This connects to what researchers call elaborative encoding: the richer the context around a word when you first learn it, the more retrieval pathways you build.

For deeper context, combining SRS with comprehensible input, reading or listening at the edge of your ability, adds the kind of natural, varied exposure that flashcards alone cannot provide. If you are curious about how the bilingual brain actually stores and switches between languages, the post on how the bilingual brain handles grammar explores this in detail.

A common mistake is treating these as competing approaches. The most effective learners typically use SRS for deliberate vocabulary building and extensive input for consolidation and naturalization.

How to Build a Spaced Repetition Habit That Lasts

The largest source of SRS failure is not the algorithm: it is consistency. Missing two or three weeks causes a card debt that discourages return. Research on how long it takes to form a habit suggests the key is reducing friction to near-zero at the start: begin with a five-minute session, a fixed trigger (morning coffee, bus commute), and a very small new-card quota of five to ten words per day.

For most adults learning a new language, research suggests:

• 10-15 minutes of review per day is sufficient to maintain an active deck of 500-2,000 words.
• Adding 10-15 new cards per day builds a 3,000-word base in about six months, enough for conversational fluency in most languages.
• More than 20-30 new cards per day typically creates unsustainable review loads within weeks.

For a broader look at language learning myths that often derail learners before they even start a review routine, the post on common myths about being "bad at languages" addresses some of the most persistent misconceptions.

Four Common SRS Mistakes That Hurt Retention

Even learners who use SRS faithfully sometimes hit a wall. These are the most common causes:

1. One-sided cards only: recognizing a word when you see it (recognition) is easier than producing it from a translation (recall). Build both directions, especially for active speaking and writing.
2. Too many new cards too fast: enthusiasm in week one becomes a 300-card review pile in week three. Steady modest quotas compound better than burst-and-crash cycles.
3. Marking cards easy when they are not: if you barely recalled a word but mark the card "easy," the interval grows too fast and builds false confidence. Honest self-assessment is the fuel of the algorithm.
4. No context on cards: bare word-pairs without a sentence, image, or audio cue are harder to encode and more likely to produce tip-of-the-tongue failure in real use. Add at least one context sentence per card.

FAQ

Source: Cepeda et al. (2008), Psychological Science; Bahrick et al. (1993), Psychological Science; Kim et al. (2022), Language Learning

About the author

Dao Huy (Lucas) is a professional translator working between English, Vietnamese, Chinese, and French, with over seven years of experience across legal, technical, and educational content. He writes these explainers out of genuine curiosity: the science of how people learn, remember, and use languages is inseparable from the daily craft of translation, where precision and cultural nuance live side by side.

If you are working on content that needs to reach Vietnamese-speaking audiences, Lucas offers English-Vietnamese translation and certified document translation services for individuals and businesses. Feel free to reach out for a free quote.