Spaced repetition: How long to wait between reviews

A phone reminder that says "Portuguese phrases, again" can do most of the work of spaced repetition. Spaced repetition means learning something once and coming back to it on later days, with a gap before each review. The right gap depends on when you'll need it: a day or two for next week, a few weeks for next year.

You may have heard it described as reviewing at increasing intervals, and growing gaps are a fine way to do it. Still, the size of each gap matters more than whether it grows. Pick a gap that fits the date you'll need it by.

Quick answer: What is spaced repetition?

• Spaced repetition is a way to learn where you review the same material on separate days, often with a longer gap each time.

• The spacing effect is the name for the result. Spread the same total study time over separate sessions, and you'll remember more than you would after one long session.

• The best gap is a share of the time until you need the material, and that share gets smaller as the time gets longer.

• When you're not sure how long to wait, wait longer. A gap that's too long costs much less than one that's too short.

• A calendar reminder is enough for a few things, and a flashcard app helps once you have hundreds of items to keep track of.

What does the forgetting curve have to do with spaced repetition?

The forgetting curve is the pattern of how you lose something you've just learned, quickly at first and then more slowly. Hermann Ebbinghaus mapped it in the 1880s, and he did it the hard way: memorizing lists of made-up syllables and timing how much less work it took to learn each list again later. Spaced repetition puts that saving to work. You come back while some of it is still there, so each review is quicker than learning from scratch.

Ebbinghaus also found the payoff of spreading the work out. In his 1885 book, 38 repetitions of a list spread over three days did as well as 68 repetitions the day before, and he concluded that spacing them out "is decidedly more advantageous than the massing of them at a single time." He tested only himself, so it was a start rather than the final word.

Later research backed him up. A 2006 review of 271 comparisons in word-recall studies found that with the same amount of study time, people remembered 47% on average after spaced study and 37% after massed study, the researchers' term for doing it all in one go.

How long should you wait between reviews?

The right wait is a slice of the time you have before you need it, and the slice shrinks as that time grows. The clearest evidence comes from a 2008 study in which 1,354 adults learned 32 trivia facts, reviewed them once after a gap of up to 105 days, and took a test as much as a year later. The best gap was about 20 to 40% of the wait when the test came a week after the review, and about 5 to 10% when it came a year after.

In the same study, recall rose as the gap grew, then slipped only gently, and the authors found the costs of waiting too long "much smaller than the costs of using too short a gap." A review that comes a few days late is fine, but one that comes too soon is the real waste.

If you want to keep something for years, stretch the gaps to a month or more. In the 2006 review, that's what did best when the test was six months or more away.

Treat the table as a rough guide. No single gap fits every goal. Most spacing studies tested memory within a day, and the two main studies above used word lists and trivia, not skills.

Do you need a spaced repetition app?

You don't need an app for a handful of things. A spaced repetition app earns its place when you have more items than you can track yourself, like hundreds of vocabulary words. It keeps a separate gap for each one and changes it as you get answers right or wrong.

Some of these apps grow the gaps by a formula. SM-2, which Piotr Wozniak described in 1990, schedules the first review a day after you learn something and the second six days later. After that, it multiplies each gap by an "ease" number that starts at 2.5, and a card you miss goes back to the start. Anki, a flashcard app, has a built-in scheduler based on SM-2. Its manual also describes a newer option you can switch on, called FSRS, which aims for 90% recall by default.

Growing gaps are a reasonable design, but they aren't required. The 2006 review found "insufficient data to say whether they help." In a 2007 vocabulary experiment, growing gaps did better on a test 10 minutes later, while equal gaps did better two days later. Equal gaps are a fair schedule, too.

Can you do spaced repetition with just a calendar?

Yes, and for one goal, it takes three reminders. Say you fly to Lisbon in five weeks and want 20 Portuguese phrases down, like ordering a coffee, asking for directions, and saying thank you.

Learn them on a Saturday, then set a calendar reminder for about a week later with the phrases in the note. At each review, cover the Portuguese, read the English, and say each phrase from memory before you check. Then go over the ones you missed.

From there, a rule of thumb does the job: Add a review about two weeks after the first and a last one a few days before you fly. If a week gets busy, move a review later, not earlier.

Once you land, the phrases are only part of what you'll need. You'll also be keeping track of the way back to your hotel from the old town and listening closely to answers in a language you're still learning. Elevate, a brain training app, has memory games for that kind of recall:

• Sequencing helps you practice recalling directions and the order of events.

• Attention helps you become a more attentive listener and exercise your short-term memory.

Elevate is built on learning principles that include spaced repetition and active retrieval, which means pulling the answer from memory instead of rereading it. Sequencing and Attention are two of its memory games, and its 40+ games also cover reading, writing, speaking, and math, with difficulty that adjusts as you improve. The memory games are part of Elevate Premium. To practice directions and listening before you fly, try Sequencing and Elevate's other memory games.

Is cramming ever enough?

Cramming beats not studying for something you need tomorrow, but it doesn't last. The authors of the 2008 study warn that squeezing learning into too short a period is likely to produce "misleadingly high levels of immediate mastery that will not survive the passage of substantial periods of time." With the same study time, the 2006 review found spacing ahead even on tests given within a minute.

Spacing tells you when to review. To make something stick in the first place, like a name or a list in order, try these other ways to train your memory.

Frequently asked questions

What is the 2-3-5-7 study rule?

The 2-3-5-7 rule is a way to plan reviews backward from an exam, described in Birmingham City University's revision guide. You study the day before the exam, then go back two more days, then three, five, and seven more, so your sessions fall 1, 3, 6, 11, and 18 days before the exam. It's a planning habit rather than a tested schedule.

What's the difference between the spacing effect and distributed practice?

The spacing effect and distributed practice describe the same idea from two sides. Distributed practice is the habit of spreading study out, and the spacing effect is the better memory that results. A 2013 review of study techniques uses one term for both, writing "we use the term distributed practice to encompass both spacing effects ... and lag effects," the second meaning the extra benefit of a longer gap over a shorter one.