What actually works, in order
Memory research is unusual in that the practically useful findings are old, robust and widely ignored — while the popular advice is mostly the opposite. Five techniques below, ordered by how much evidence supports them.
One thing to get out of the way first: re-reading is nearly useless, and it is what most people do. It produces a strong feeling of familiarity that is easily mistaken for knowing something. That mismatch — feeling fluent while being unable to recall — is the single most common study mistake there is.
1. Retrieval practice — testing yourself
The best-supported technique in the entire literature, and the least popular. Trying to recall something strengthens the memory more than reviewing it does.
The classic demonstration: groups given the same total study time either re-read material repeatedly or read once and then tested themselves. On a delayed test, the self-testing group performed substantially better — despite feeling less confident at the time.
- Close the book and write down what you remember. Then check. The gap you find is the actual learning.
- The struggle is the mechanism. A retrieval that comes easily strengthens little; effortful recall strengthens a lot. Difficulty is not a sign it is failing.
- Wrong answers still help, provided you get feedback. Guessing then correcting beats not attempting.
2. Spaced repetition — spreading it out
The same total study time produces far more durable memory when distributed across days than when massed into one session. Known since Ebbinghaus in the 1880s and confirmed continuously since.
- Review just as you start to forget. Reviewing something you still know well is wasted effort; the gain is largest when recall is effortful but possible.
- Expanding intervals. Roughly: one day, three days, a week, two weeks, a month.
- Cramming is not useless — it is temporary. It works for a test tomorrow and fails for a test in a month. Choose deliberately.
Combining this with retrieval practice — spaced self-testing — is the single most effective study method known, which is why every serious flashcard app implements exactly that.
3. Chunking — making units bigger
Working memory holds roughly four items. The way around the limit is not to hold more items but to make each item bigger.
A phone number as ten digits is impossible; as three groups it is easy. A chess position is meaningless to a novice and a single recognisable pattern to a master — which is why masters have extraordinary recall for real positions and ordinary recall for random ones. Their memory is not better; their units are.
- Find structure before memorising. Understanding how something is organised turns a list into a shape.
- Give chunks names. A named group survives a delay much better than an unnamed one.
You can watch chunking work directly on our flash memory task — people who see the lit cells as a path rather than as separate positions score far higher.
4. Elaborative encoding — connecting to what you know
Memory is associative. Isolated facts have nothing to attach to and decay; connected facts are reachable by many routes.
- Ask "why is this true?" Explaining a fact to yourself binds it to what you already know. This is elaborative interrogation, and it is well supported.
- Explain it to someone else — or pretend to. Teaching forces you to find the structure, and you discover your gaps immediately.
- Use dual coding. Verbal plus visual encoding beats either alone. Saying "star, top-left" while looking at it is measurably better than only looking.
5. The method of loci — the memory palace
The technique competitive memorisers actually use. Attach each item to a location along a route you know well — your home, your commute — then walk the route to recall them.
It works because spatial memory is unusually strong and because the route imposes an order for free. It genuinely produces extraordinary results for ordered lists.
Two honest caveats: it takes real effort to set up, and it is best suited to ordered lists of discrete items. For understanding a concept, elaborative encoding and retrieval practice are more useful.
What does not work as advertised
- Highlighting and underlining. Consistently among the least effective techniques studied. It feels productive and mostly marks text without engaging with it.
- Re-reading. As above. Fluency is not knowledge.
- Learning styles. The idea that you are a "visual learner" who should receive visual material has been tested repeatedly and matching instruction to claimed style does not improve outcomes.
- Brain training games. They improve the games. See does brain training actually work?
- Supplements. With the exception of correcting an actual deficiency, the evidence for memory supplements in healthy adults is poor.
The thing that outperforms all of it
Sleep. Memory consolidation happens during sleep — this is not a metaphor, it is a well-mapped physiological process. Learning something and then sleeping produces better retention than learning it and staying awake the same number of hours.
Which means an all-nighter before an exam is doubly bad: it removes the consolidation you need and degrades the attention you will use to sit the test. If you have to choose between one more hour of review and one more hour of sleep, the evidence favours sleep.
For how attention limits what reaches memory in the first place, see how to improve your focus.