Spaced repetition: why spacing makes learning stick

The forgetting curve

In the nineteenth century Hermann Ebbinghaus described how fast new information fades if you do nothing with it. The exact shape depends on the material and the person, but the pattern is well known: much is lost quickly, the rest fades more slowly. What you hear in a one-day training is largely gone a week later.

More important is what Ebbinghaus also showed: a repetition slows forgetting. And each next repetition slows it further. That is the basis of spaced repetition.

Spacing works better than cramming

A frequently found result in memory research is the spacing effect: the same amount of practice time yields more if you divide it over more moments than if you do it in one block. Cramming works for tomorrow, spacing for a month from now.

It combines with the testing effect: if during each repetition you have to retrieve something instead of rereading, that strengthens the trace even more. See microlearning with swipeslides.

How to set it up in a learning journey

1. Divide the material into small learning moments. One idea per slide, a few minutes at a time.
2. Plan repetitions. A short learning moment after a day, after a week, after a month. The calendar helps plan the moments and a workflow can assign slides to people.
3. Use variation. Ask the same question in a different way: first multiple choice, later an open question, then in a case. In CoVince every participant can get their own variant of a question.
4. See what sticks. Questions that keep going wrong come back more often.
5. Close with application. The last repetition is an assignment in real work.

Which parts help?

Swipeslides are the natural form for repetitions: short, on the phone, with a question. The learning journey remembers where you were, so you can pick up whatever fits in between. And in a Control Room you see which group has done which step.

You can play a scenario game again after a while: a second round shows what has stuck. See scenario games.

Pitfalls

• Repeating too much. Repetition is not a punishment. Let go of what someone has mastered.
• Same question, same answers. People remember the answer to the question, not the material. Vary.
• No role for the work. Repetition in the learning journey without application stays theory. See transfer to the shop floor.

An example of a repetition schedule

Take a learning journey about a new product. Day 0: a ten-minute learning moment with the core and five questions. Day 1: three questions back, worded differently. Day 7: a short case in which the product is applied. Day 30: an open question and an assignment at work, for example explaining it to a colleague.

Each repetition is short, asks something of the learner and leaves a slightly longer gap. The schedule is an example, not a law: the rhythm depends on how important the material is and how soon people need it. At the last repetition, measure what has stuck and adjust the rhythm.

Common misconceptions

• "Repeating is boring." Only if it is the same question. With variation and application it feels like something new.
• "It has to be an app with an algorithm." An algorithm can help, but a simple schedule of three moments already yields a lot.
• "More repetition is always better." What someone has mastered does not need more. Aim the repetition at what is still shaky.

See also designing the complete learning journey.

What do you make suitable for repetition?

Not everything needs repeating. Repetition pays off for what someone must be able to do lastingly and what fades quickly: procedures, terms, rules, conversation structures. It pays off less for what can always be looked up, such as a phone number. A good question to ask: if someone needs this in three months, do they have to know it themselves or can they look it up?

• Repeat key concepts and decision rules.
• Repeat skills with a short exercise, not only with a question.
• Let go of details that can be looked up, or give a quick reference.

Make every repetition a small application: a short case, a choice, a step in order. Then someone practises not only the words, but the use.

Seeing whether it sticks

The beauty of repeating with questions is that it measures at once. A question that still goes well after a week is mastered. A question that keeps going wrong points to a weak spot in the learning journey or the material. Look per theme at how the scores develop: if they drop fast after the first learning moment, the first explanation is probably too thin or too fast.

Use the reports to adjust the schedule. A theme where people still score well after a month can get a longer gap. A theme where it wobbles gets an extra repetition. See measuring learning.

Start small

You do not have to build a system with algorithms. Start with one learning journey and three repetition moments: one day, one week, one month. At the last repetition see how much has stuck, and adjust the rhythm. That way, with a simple start, you make a learning journey that makes something last.