Spaced repetition: 130 years of one finding

Hermann Ebbinghaus, sitting alone in a Berlin study in the 1880s, memorized lists of nonsense syllables — bok, zif, vag — and tested himself at varying intervals. His 1885 monograph established the forgetting curve: a steep initial drop in retention followed by a slower decay. He also noticed that reviewing material at increasing intervals dramatically improved long-term retention.

This is the empirical foundation of spaced repetition. The finding has been replicated thousands of times, across age groups, content domains, and methodologies. It is one of the most robust findings in psychology — and one of the most widely ignored in mainstream education.

1. The basic effect

When you study material once and review it once an hour later, you remember much less a week later than if you study it once, review the next day, then the week after, then a month later. Same total study time. Different distribution. Roughly doubled retention.

Cepeda and colleagues' 2008 meta-analysis aggregated 184 studies of spacing effects. The effect was reliable across virtually every paradigm: spaced practice outperforms massed practice for long-term retention, often by 2x or more (Cepeda et al., 2008).

2. The optimal spacing

The optimal interval depends on the desired retention period. A useful rule from Cepeda's work: review at roughly 10-20% of the time you want to remember the material. To remember something for a year, review at ~1-2 months. To remember for a week, review at ~1 day. The exact ratio is task-dependent but the order of magnitude is consistent.

Apps like Anki, SuperMemo, and various language-learning platforms implement adaptive spacing algorithms that schedule reviews based on individual response history. The algorithmic approach outperforms fixed schedules and is approximately optimal for typical retention goals (Wozniak, 1990).

3. The neural mechanism

The leading hypothesis: spacing forces the brain to retrieve the memory at the point where it's becoming inaccessible. Each retrieval strengthens the consolidation pathway. Massed practice doesn't engage retrieval — the material is still active in working memory, so the brain doesn't do the reconsolidation work that produces durable long-term storage.

This is also why testing yourself outperforms re-reading. Testing requires retrieval; re-reading provides recognition. Retrieval drives consolidation; recognition mostly doesn't (Karpicke & Roediger, 2008).

4. The implication

For anything you want to remember long-term — vocabulary, concepts, names, procedures — the structure of practice matters more than total time. Three twenty-minute sessions across a week, with retrieval, beat a single hour of cramming. By a lot.

Most schools and most adult learners do the opposite. Spaced repetition is one of the easiest performance wins in self-directed learning. It just feels worse than cramming, because you forget things between sessions. That feeling is the cost of the long-term gain.