The evidence for spaced repetition

The effectiveness of spaced repetition for remembering facts is one of the best-supported results in psychology and education. The method combines two study techniques: self-testing (as opposed to rereading), and taking time out between practice sessions (as opposed to cramming).

Spacing out short practice sessions beats cramming

We’ve known about this method for a long, long time, and we have a tonne of evidence for it. The spacing technique was first proposed by Hermann Ebbinghaus, who measured its effectiveness on himself in 1885. The body of evidence grew from there, and in 2006, Cepeda et al ran a meta-analysis of 317 different experiments which showed that it continues to be a powerful technique.

The effect holds in lab settings, and in classrooms; in 2025, Mawson and Kang found it held across 22 classroom studies with more than 3,000 learners.

The effect lasts for years. In a nine-year study, four people learned 300 foreign-language word pairs, and relearned them at gaps of 14, 28 or 56 days. When they were tested (up to five years later!), 13 sessions spaced 56 days apart had worked about as well as 26 sessions spaced 14 days apart.

Flashcards work. Kornell (2009) asked participants to learn word pairs from digital flashcards, either spaced out or crammed. Spacing had worked better for 90% of them, but after the first session 72% believed that cramming had worked better.

Spacing has also been tested in medical training. In a randomised controlled trial, medical students were emailed a question per week on two of the four urology topics that they had just studied. At the end of the year, they scored higher on the topics they’d kept being asked about.

Using longer spaces is well supported. A 2008 study taught more than 1,350 people a set of facts, gave them a second session up to three and a half months later, and tested them up to a year after that. To remember something for a week, the best gap was between one and three days. To remember it for a year, it was three to five weeks.

Based on that evidence, when you mark a card as ‘remembered’, Flora will grow the space between that review and the next one. After a few successful reviews, a card can stay deep in the deck for weeks without needing to be reviewed again. After a few months of daily practice, everything you want to remember can be committed to your long-term memory.

Self-testing beats rereading

In 2006, Roediger and Karpicke gave students short passages to learn. One group of students read their passage four times in quick succession. Another group read their text once, then wrote down what they could remember three times. Immediately after the learning session, the rereading group could remember more of the text. But one week later, the students who had tested themselves remembered 61% of the passage, and the rereaders only remembered 40%. Self-testing leads to a deeper embedding of concepts into your long-term memory.

The same thing has been observed in the real world, out of laboratory conditions. Yang et al (2021) combined studies of more than 200 classrooms, covering more than 48,000 students, and found that being quizzed raised students’ achievement. In a randomised trial with trainee doctors, residents were given a review sheet to study for one topic, and repeatedly tested on the other. Six months later, they scored 26% on the one they’d studied, and 39% on the topic which they had been tested on.

When Dunlosky et al rated 10 common study techniques, self-testing and spacing out practice sessions were the only two techniques which they rated as highly useful. The flashcard system which Flora is based on uses both techniques: you ask yourself a question before revealing the answer (self-testing), and the same question is repeated after longer and longer gaps (spacing).

Making flashcards takes time

We know that most students who use flashcards would rather not write their questions and answers themselves, and that the biggest hurdle to learning is getting started at all. In a survey of 901 undergraduates at a US university, 56% of students preferred premade flashcards to their own, commonly citing convenience and time saving. Among the students at one US medical school, at least 80% of participants using a spaced repetition system were using decks pre-made by someone else, and 57% made cards of their own from lectures (Lu et al, 2021).

Based on this, Flora jump-starts your learning by creating a brand new, complete deck of flashcards in seconds, for any subject. Getting started is the hardest part of learning something new, but Flora makes it easier.

However, we also know that making cards isn’t fully wasted time. There’s a real skill to writing good flashcards (which we’ve tried to capture in our card-generating pipeline). The best decks are personal, based on time spent thinking critically about the text and what you know or don’t know. When Pan et al (2023) gave people the same amount of time either to make their own flashcards or to use premade ones, the people who made their own tended to remember more after two days.

With that in mind, Flora still gives you the ability to write new cards, and to edit or delete any which appear in an automatically generated deck. It’s encouraged - taking ownership of your decks and tuning them to your needs is a very important part of the puzzle.

What the research doesn’t show

Flora won’t make you an expert in real-world situations on its own. Daily practice can help you build a solid mental skeleton of facts, vocabulary, definitions, etc, but it’s up to you to fill in the gaps and make them useful in your day-to-day life.