The Best Study Techniques Backed by Science

The highlighter problem

Walk through any university library in exam season and you will see the same scene repeated at every table: a textbook open, a yellow marker in hand, three-quarters of the page already glowing. It looks like work. It feels like work. The pen moves, the page changes colour, the hours pass. And when the exam arrives, the material turns out to be somewhere between vaguely familiar and completely gone.

The uncomfortable finding from decades of research on study methods is that the techniques students use most are the ones with the weakest support. Highlighting, rereading, and summarising as you go all rate poorly in reviews of learning strategies, not because they do nothing, but because they do very little for the time they consume. The University of North Carolina's learning centre puts the same point bluntly in its own study advice: recognising information on a page is not the same as being able to produce it.

There is a decent reason the illusion is so persistent. Rereading makes text easier to process, and we read that ease as knowledge. Psychologists call it fluency. Your brain notices that the sentence went down smoothly and concludes, wrongly, that the idea is now yours. Difficulty, not ease, is the better signal.

Test yourself before you are ready

The single most reliable upgrade available to most students is to stop asking whether they recognise the material and start asking whether they can retrieve it. Close the book. Write down everything you can remember about glycolysis. Compare, correct, repeat tomorrow. It is unpleasant in a specific way — the blank page is a small humiliation — and that discomfort is the mechanism working.

A review in the Yale Journal of Biology and Medicine aimed at medical students reduced effective learning to three instructions: test yourself, learn actively, repeat in intervals. Retrieval does more than measure what you know. Research published in Frontiers in Psychology on the forward effect of testing found that being tested on earlier material improves how well people learn material that comes afterwards, which means self-quizzing is not merely a check on progress but part of the studying itself.

Practically, this means turning your notes into questions the day you take them. Our biology student, revising a metabolism unit, should not be rereading her lecture slides. She should be sketching the pathway from memory on a blank sheet, naming the enzymes she cannot recall, then checking the slides only to repair what she got wrong. Fifteen minutes of that beats an hour of scrolling through her own highlighted PDF.

Spread the work across days

The second technique with strong support costs nothing but planning. Four one-hour sessions on four days beat one four-hour session on the night before, and the gap is not small. The Institute of Education Sciences practice guide on organising instruction and study to improve student learning opens with exactly this recommendation, because the evidence for distributed practice has been stable for a century.

Cramming does work, briefly. If the exam is tomorrow morning, cram. The problem is that material learned in one dense block decays fast, which is why students who crammed through a first-year course often arrive in second year convinced they never learned it — they didn't, not durably. Spacing costs the same total hours and leaves you with something in June.

The scheduling does not need to be elegant. Touch each topic three or four times across a fortnight, with at least a day between visits, and let the review be short: ten minutes of recall, not a full re-study. Our metabolism unit gets revisited on Tuesday, Friday, and the following Wednesday, each time as a set of questions rather than a reread.

Mix your problems instead of blocking them

Textbooks are organised by chapter, so students practise by chapter: twenty problems on one method, then twenty on the next. Performance during that kind of blocked practice looks excellent. You have primed yourself for one procedure and you execute it well. On the exam, where the questions arrive unlabelled, the hard part turns out to be choosing the method, and blocked practice never trained that.

Interleaving mixes problem types within a session. It feels ragged and error-prone, and it produces better performance later, particularly in maths and any subject where similar-looking questions require different approaches. For our biology student, interleaving means a practice set that jumps between glycolysis, the electron transport chain, and photosynthesis rather than grinding through each in isolation.

There is a cost worth naming: interleaved practice makes your study sessions feel less successful. Fewer questions answered smoothly, more hesitation. Students routinely rate blocked practice as more effective while performing worse on delayed tests. Trusting the feeling here is a mistake.

Explain it until the gaps show

Elaboration means asking why something is true and how it connects to what you already know, out loud or in writing, until you hit the point where the explanation collapses. That collapse is the useful part. It shows you the exact edge of your understanding, which reading never does because a textbook explanation always sounds complete.

Self-explanation and teaching a topic to someone else both work through the same route. The person you teach does not need to exist; an empty room and a whiteboard will do, though a patient friend who asks 'why?' is better. The American Psychological Association's overview of learning and memory describes memory as reconstructive rather than a recording, and explanation is the cheapest available way to test whether the reconstruction actually holds.

What the techniques cost and return

Ranking study methods only makes sense against effort. A technique with a modest effect that takes five minutes may be a better buy than a powerful one that requires an hour of setup. The table below sorts the main options that way, which is roughly how you should choose between them when the exam is three weeks out and you have four subjects.

Building a week that actually works

A workable revision week does not need to include every technique. Take one topic, generate questions from your notes on the day you make them, quiz yourself on that topic three times over the next ten days, mix in problems from adjacent topics each time, and once per topic try to explain the whole thing from memory to an imaginary sceptic. That is the entire system, and it fits inside the hours you already spend.

The failure mode is drifting back to rereading whenever the recall attempts get uncomfortable, usually in the first week when your retrieval is poor and the textbook is right there, sounding reassuring. Expect the discomfort, treat it as the price rather than a signal that the method is wrong, and give it two weeks before judging. The evidence has been consistent for a long time; the difficulty is entirely one of habit, which is a separate skill worth building on its own terms.