All posts Learning science

How spaced repetition actually works, and why cramming doesn't.

A crammed night before a test feels productive. The memory science says it mostly isn't — and explains exactly what to do instead.

Published 20 Sep 2026 · 6 minute read

Every parent has seen it: a child re-reads their notes the night before a test, does fine the next day, and can barely explain the same topic a week later. That's not a discipline problem. It's what memory researchers call the forgetting curve — a pattern first documented by psychologist Hermann Ebbinghaus in the 1880s and reproduced in a well-known modern replication, which found that without any review, most newly learned material fades sharply within the first few days (Murre & Dros, 2015, PLOS ONE, accessed 2026-09-20). Cramming works against this curve, not with it.

Why spacing beats cramming

The fix researchers keep landing on is the spacing effect: spreading the same material across several separate sessions, rather than one long block, produces better long-term retention. A well-cited meta-analysis pooling 63 studies found a real, moderate benefit to spaced over massed practice (Cepeda et al., 2006, accessed 2026-09-20). A follow-up study went further and found something genuinely useful for planning study time: the ideal gap between reviews isn't fixed, it grows as the material gets older, and it depends on how long you actually need to remember it for — a test in two weeks calls for a different review schedule than one in six months (Cepeda et al., 2008, Psychological Science, accessed 2026-09-20).

One catch worth knowing: the benefit of spacing usually shows up on delayed tests, days or weeks later, not on a quiz taken right after the study session. That's exactly why cramming can feel like it's working — same-day performance often looks fine either way. The gap only opens up once time has passed, which is precisely when it counts for an actual exam.

The second piece: testing yourself, not just re-reading

Spacing tells you when to review. The other half of the puzzle is how. Simply re-reading notes is one of the least effective ways to study, because it creates a false sense of familiarity without exercising recall. The better approach — called retrieval practice or the testing effect — is to actively try to produce the answer from memory, then check it. A meta-analysis of dozens of studies found retrieval practice consistently outperformed restudying, with a real and fairly large effect on later recall (Adesope, Trevisan & Sundararajan, 2017, accessed 2026-09-20). Put simply: a short quiz on yesterday's topic beats reading the same page over again, even though re-reading feels easier in the moment.

Spacing and retrieval practice work best together: a short quiz on material from a few days ago, repeated at gradually widening intervals, beats either technique alone. This is genuinely one of the best-replicated findings in learning science — not a trend, a decades-old result that keeps holding up.

What this looks like in an ordinary week

  • Little and often beats long and rare. Four 15-minute sessions across a week generally beat one 60-minute session the night before.
  • Quiz before you re-read. Try to recall the answer first — even getting it wrong and then checking helps more than reading the answer cold.
  • Let old topics resurface. Don't retire a topic once it's "done." A quick check-in a week later, then a month later, is what actually locks it in.
  • Expect review to feel a bit harder than re-reading. That's not a bad sign — the difficulty is exactly what makes it stick (researchers call this "desirable difficulty").

None of this requires special software — a stack of flashcards and a calendar reminder will do it. It's exactly why this is one of the cheapest, highest-leverage habits a family can build, and why so many effective study tools are just a scheduling system wrapped around these two ideas.

This is exactly why SparkIQ's tutor schedules review at 1, 3, 7, 16 and 30 days automatically, and quizzes with mixed retrieval questions rather than letting a child pick "review" and just re-read. It's additional support alongside school and teachers, not a replacement for either — try it free, no account needed to start.