How to study: what works, and what only feels like it works

Two methods survive the evidence

Strip away the study-tips industry and the research record is almost embarrassingly simple. Across hundreds of experiments, spanning decades, two techniques keep winning: testing yourself on material instead of re-exposing yourself to it, and spacing that testing out over days and weeks instead of massing it into one sitting. Psychologists call them retrieval practice and distributed practice. Everything else, from highlighting to rereading to recopying notes, sits somewhere between marginal and near-useless in comparison.

The effect sizes are not subtle. Students who spend a third of their time reading and two thirds testing themselves reliably outperform students who read the whole time, on the same material, with the same hours. Spacing multiplies the effect: the same number of practice attempts, spread across a week, can double what survives to exam day compared with the identical attempts made in one evening.

So the honest answer to “how should I study?” is short: make yourself produce the material from memory, repeatedly, with growing gaps between attempts. The rest of this guide is about why almost nobody does that, and how to set your studying up so that you actually do.

Why the popular methods feel better than they work

Rereading is the most used study method in every survey of students, and one of the worst performers in every controlled comparison. The reason it persists is that it feels excellent. The second pass through a chapter flows; recognition kicks in; the material seems obvious. That fluency is real, but it is the wrong signal. The exam will not ask whether the page looks familiar. It will ask you to produce the idea from nothing, and production is precisely the thing rereading never rehearses.

Researchers call this the illusion of competence, and it explains most bad studying in one stroke. Highlighting feels like engagement but is usually just decorated rereading. Recopying notes feels like work because it is work, just not the kind that builds recall. Watching a worked video example feels like understanding, and is, right up until you try to do one alone. None of these are evil; they are simply preparation for the real work, mistaken for the work itself.

The flip side is that the methods that do work feel bad. Testing yourself before you feel ready is uncomfortable. Retrieving something three days after you learned it, when it has gone halfway foggy, feels like evidence of failure. It is the opposite: the difficulty is the mechanism. Memory strengthens in proportion to the effort of the retrieval, which means the studying that works is recognisable by a mild, persistent strain that the popular methods are designed to avoid.

Testing yourself, without writing tests

Retrieval practice does not require anyone to hand you a quiz. Close the book and write out what you remember. Explain the idea aloud as if teaching it. Answer the end-of-chapter questions before rereading the chapter. Any act that forces production counts, and the crudest self-test beats the most elegant summary. The full case, including why flashcards are both the most popular and the most abandoned implementation, is in our guide to active recall.

Spacing, without a calendar

The second pillar is timing. Memory decays on a curve, and the most efficient moment to revisit anything is just before it slips away: too soon wastes the attempt, too late means relearning. Managed by hand, this becomes a scheduling problem across dozens of topics, which is why most people who try it with spreadsheets or card apps quietly stop. How the schedule can run itself, especially for subjects where reviews should be fresh problems rather than flashcards, is covered in spaced repetition without making a single card.

Studying against a deadline

All of the above assumes time. With an exam in sight the question changes shape: the night before, the week before, and the month before reward different moves, and pretending otherwise is how study advice loses people. We keep the honest version, including what cramming genuinely can and cannot buy you, in how to study for a test.

Maths is its own case

Everything here applies to maths, and none of it is sufficient, because maths adds two constraints most subjects do not have. The unit of study is the solved problem, not the reread page, and the subject is ruthlessly cumulative: a wobble in fractions surfaces years later as a wall in algebra. Studying it well means treating it as a skill to practise, not a subject to revise, and sometimes it means walking back down the tree first. If you are rebuilding from a long way back, learning maths from scratch covers the return journey, and if focus itself is the constraint, studying maths with ADHD takes the systems angle further.

The vocabulary of learning itself

If you want the conceptual layer underneath these methods, the ideas made famous by the learning how to learn tradition: focused and diffuse thinking, chunking, and the illusions of competence above, we have distilled them into a separate short guide. It is worth an afternoon; the concepts make the methods make sense.

Where to start

Not with a schedule template. The highest-leverage first step is an honest inventory of what you can currently produce from memory, because every method above operates on material, and the material worth studying is defined by where your actual gaps are. Map that first, against a map of every concept and what it stands on if maths is your subject, and let the methods do their work from an honest starting point.