Ten revision techniques ranked by evidence

In 2013 a team of cognitive psychologists did something unusually useful: they took ten study techniques students actually use, reviewed the evidence for each, and graded them. Two came out high. The two most children default to came out low.

At a glance

The paper

Dunlosky, Rawson, Marsh, Nathan and Willingham, Improving Students' Learning With Effective Learning Techniques: Promising Directions From Cognitive and Educational Psychology, published in Psychological Science in the Public Interest in 2013.

It is long, freely discussed, and written to be read by people who teach rather than only by researchers. The authors graded each technique on how well it generalises — across materials, ages, and the kinds of tests that matter — and were explicit that a low rating often meant "insufficient evidence in realistic conditions" rather than "proven useless".

If you read one piece of education research as a parent, this is a strong candidate.

The two winners

Practice testing. Retrieving material from memory rather than reviewing it. Rated high because the effect is large, appears across ages, works with a wide range of material, and shows up in real classrooms rather than only in laboratories. This is the finding flashcards exist to exploit.

Distributed practice. Spreading the same total study time out rather than massing it. Rated high for the same reasons — it is robust, it is well understood, and it costs nothing to implement.

Notice what these two have in common: neither requires new material, a new app, or extra hours. They are both about rearranging effort a child is already spending. That is what makes them such good advice.

Notice also that they are the two mechanisms a spaced-repetition flashcard system combines by construction. That is not a coincidence — it is why this category of tool exists.

The two most children actually use

Rereading. Rated low. It is the default study behaviour almost everywhere, and its problem is that it produces fluency rather than knowledge. The second pass through a page feels much easier than the first, and that ease is misread as learning.

Highlighting and underlining. Also low. Worse than rereading in one respect: it produces a visible artefact — a colourful page — that feels like evidence of work. The review found little benefit, and noted a plausible cost: attention goes to marking rather than to thinking about meaning.

If your child's revision consists of reading notes with a highlighter, they are doing the two lowest-rated techniques simultaneously, and it will feel more productive than the thing that would actually work.

The middle: better than their rating suggests

Three techniques landed at moderate, and each is worth knowing.

Interleaved practice — mixing topics within a session instead of blocking one topic at a time. The evidence has strengthened since 2013, particularly in maths, where mixing problem types forces students to work out which method applies rather than mechanically applying the one from the top of the page. It is why our review queue deliberately mixes decks rather than serving one lesson's cards in a block.

Elaborative interrogation — asking why a fact is true rather than only what it is.

Self-explanation — explaining your reasoning as you work through a problem.

Both of these are essentially what a Socratic tutor does, and both are hard for a child to do alone, which is a fair summary of why tutoring exists.

A note on learning styles

Not in the ten, but worth raising because it dominates conversations about how children learn.

The idea that a child is a "visual learner" or an "auditory learner", and should be taught in that channel, has repeatedly failed to hold up. Studies that test the actual claim — matching instruction to a stated preference should improve outcomes — generally find no benefit. Children do have preferences; the preferences do not predict which teaching works better.

This matters here because it is the most common way a study product oversells personalisation. Adapting tone, pace and examples to a child is real and useful. Claiming to adapt the method to a learning style is not, and we do not.

What we did with this

Fairly directly:

• The daily loop is practice testing, spaced. The two high-utility techniques, made the default rather than offered as an option.
• The review queue interleaves across decks and shuffles within them, so consecutive cards are not about the same idea.
• The tutor does the elaborative interrogation and self-explanation — asking why, asking what your child notices — because those are the moderate-rated techniques that need another person.
• Nothing in the product encourages rereading or highlighting. There is no highlighter, and notes are not the interface.

The uncomfortable implication of this literature is that the most popular study behaviours are popular because they feel good, and the effective ones are unpopular because they feel like work. A product built on the evidence has to be willing to feel slightly harder than the alternative — while making sure the difficulty is the useful kind.

FAQ

Which study techniques have the strongest evidence?

Practice testing and distributed practice. In Dunlosky and colleagues' 2013 review of ten techniques, those were the only two rated high utility.

Does highlighting help?

The review rated highlighting and underlining low utility. They feel productive, take almost no effort, and produce little durable learning — which is a bad combination because the feeling is what students trust.

What about learning styles?

Matching teaching to a supposed visual or auditory learning style has repeatedly failed to show benefits in controlled studies. It is one of the most persistent myths in education.