Study Techniques Ranked by the Evidence Behind Them

Here is the ranking, the numbers behind it, and where it stops applying.

The ranking, and who produced it

The widest direct comparison is Dunlosky and colleagues' review of ten common techniques, built for exactly this question: which habits deserve a student's hours. Two techniques landed in the highest utility group, practice testing and distributed practice. The two most popular ones, rereading and highlighting, landed at the bottom. The striking part is not that a gap exists but how the two lists relate: what most students do most is what the comparison ranks lowest. Nothing in that ranking says rereading does zero work. It says that an hour spent rereading holds less by exam day than the same hour spent another way.

Self-testing: the largest measured gap

Closing the book and producing the answer feels slower than reading the page again, and it wins by a wide margin. In Roediger and Karpicke's experiments, students who read a passage once and then practised recalling it outperformed students who reread it repeatedly when everyone was tested days later. The follow up in Science put numbers on the delayed test: close to eighty percent recall for the retrieval group against roughly a third for the rereaders. The rereaders had also predicted they would do just fine, which is the part worth keeping: how studying feels is a poor witness to what it leaves behind.

Spacing: the gap scales with your horizon

Three hours on one topic in one sitting and the same three hours across three days do not leave the same trace. Cepeda and colleagues measured the useful gap between reviews across many delays and found it is not a fixed number of days. It scales with the retention horizon, landing roughly between a tenth and a fifth of the time until you need the material. A test in a week puts the first return about a day out; a final in June moves it to weeks. The full arithmetic lives in the review interval guide.

The first day decides the most

The reason early returns matter is the shape of forgetting. Murre and Dros reran Ebbinghaus's original experiment and the curve came back unchanged: the steepest loss arrives within the first day, with roughly two thirds of new material gone by its end, and a much slower slide after. A ten minute closed book recap on the evening of the lecture protects more than an hour pushed to the weekend, because it lands where the curve is steepest.

Interleaving: slower practice, stronger exams

Twenty problems of the same type in a row feel fluent. Mixing types feels rougher and resembles the exam more closely, because interleaved practice trains a skill blocked practice never touches: deciding which method a problem calls for. The exam exercises that decision on every question, and the student who mixed topics has already made it hundreds of times.

Why the low ranked techniques survive

If rereading measures so poorly, why is it everywhere? Because it is easy to start, easy to continue and generous with feedback of the wrong kind. Every pass makes the page more familiar, and familiarity is read as progress. That signal is real; it is just about recognition rather than production. Self-testing inverts the experience: it feels worse precisely when it works, because the struggle to produce an answer is the exercise itself. A technique that punishes you while it pays and a technique that flatters you while it does not are hard to tell apart from the inside. That is what the delayed test is for.

What this ranking does not settle

Three limits. The studies behind these numbers used defined material tested at set delays, so treat the figures as directions rather than targets for your own syllabus. The ranking describes averages; a given course and a given student can sit far from them. And no technique on the list produces understanding. They keep understood material reachable. First contact still belongs to reading, lectures and asking questions, and a page you never understood will not be rescued by any schedule of returns.

Putting the top two to work

The honest obstacle is that both winning techniques demand preparation: questions to test with, a calendar of returns. That preparation is the step most students skip, and it is the step that automates well. Quiz generation turns your own notes into gradeable questions, spaced repetition runs the return schedule, and the head to head comparison walks through the retrieval evidence in detail. For the timing half, start with how long to wait between reviews.

By Mert Erarslan (Exam prep and subject routes)

What is the most effective study technique?

In direct comparisons, practice testing and distributed practice sit in the highest utility group. Combined, the same hours leave measurably more behind.

Is rereading useless?

Not useless, just weak per hour. The delayed test comparison put retrieval practice near eighty percent recall against roughly a third for repeated reading.

Does highlighting help at all?

Very little on its own. It can support a later step, like turning marked lines into questions, but as a standalone technique it ranks at the bottom.

When should the first review happen?

On the day you learned the material. Roughly two thirds of new material is gone by the end of the first day, so an early return protects the most.

Why does mixed practice feel harder?

Because it adds a decision blocked practice removes: which method does this problem need. That difficulty is the training, and the exam asks for it on every question.

Do these findings apply to problem solving courses?

The strongest numbers come from recall of defined material. Retrieval and spacing appear to help with applied work too, but the measured effect sizes there are not established by these studies.

Last updated: 2026-08-18 · Written with AI assistance and reviewed before publishing.

Study techniques ranked by evidence: the comparison most study advice skips