Why Revising One Topic at a Time Backfires
In two studies involving nearly 1,000 students, researchers found that mixing up topics is far more effective than focusing on just one at a time. Even when given the exact same practice questions, students who mixed their topics scored 23 to 34 percentage points higher.
Your child spends an hour on quadratics. By the end they are getting every question right. They close the book feeling like they have cracked it.
Three weeks later the mock comes back and the quadratic question has cost them four marks.
The usual conclusion is that they need more practice. Often what they need is differently ordered practice, and the research on this is stronger than almost anything else in education.
Everything below comes from two randomised trials on pupils of about twelve, working on early algebra. That is younger than most of the students we teach, and I will come back to what that does and does not let you conclude.
The two studies
In the larger one, published in 2020, 787 pupils across 54 classes in five schools were randomly assigned to one of two kinds of homework for a term. One group worked through topics in blocks, the way most textbooks are arranged. The other group got the same problems, the same number of them, shuffled so that consecutive questions came from different topics.
About six weeks later they sat an unannounced test.
The mixed group scored 61%. The blocked group scored 38%.
The earlier study, from 2014, used four topics that had nothing to do with each other: solving linear equations, proportion word problems, graphing equations and calculating slope. Pupils practised for nine weeks, then sat an unannounced test two weeks later.
72% against 38%.
Both gaps are large by the standards of classroom research, where an intervention that moves anything at all is considered a success.
What “mixed” actually means here
Not revising several subjects in one evening. This is all within maths.
It means a worksheet where question 1 is a proportion problem, question 2 asks for the gradient of a line and question 3 is an equation to solve, rather than twelve proportion problems in a row.
The part worth pausing on: both groups did the same problems. Same questions, same quantity, same difficulty. The only thing that changed was the order they came in.
So this is not a matter of working harder or longer. It is free.
Why it works, and it is not what you would guess
The obvious explanation is that mixing stops topics blurring into one another. The 2014 study checked, and it was not that. Only 4.4% of wrong answers involved a pupil reaching for the wrong method.
What mixing trains is the step before the method: looking at a question and working out what kind of question it is.
A topic worksheet does that step for your child. The heading says “Quadratics”, so every question is a quadratic, and the decision is made before they pick up the pen. They practise the second half of the skill, the execution, and never touch the first half.
An exam paper never tells you. Question 18 sits there with no label on it.
That is the whole gap in one sentence: a worksheet tells you the method, and an exam does not.
The part that makes it hard to stick to
Mixed practice feels worse while you are doing it. More questions come out wrong. There is no run of ticks and no moment of having got it.
That feeling is exactly the trap. It measures how the practice session is going. It says nothing about what will still be there in three weeks, and the studies show the two come apart.
The researchers also note that the advantage shrinks when the test comes soon afterwards. If the test is tomorrow morning, an hour on one topic may genuinely look better. It is at a delay of weeks, which is the real distance between revision and an exam, that the gap opens up.
Where this does not apply
Three limits worth being straight about.
These were twelve year olds doing early algebra. Applying the finding to GCSE and A-Level is reasonable, and it matches what we see, but it is an extrapolation rather than something those trials measured.
You cannot mix what has not been taught yet. The authors are explicit that pupils probably need blocked practice when a method is new. Block it to learn it, mix it to keep it. The mistake is not starting with a topic, it is never leaving.
Feedback has to come with it. Mixed practice where nobody finds out which answers were wrong is just a more efficient way of practising mistakes.
What to do with it
Once a method has actually been taught, stop letting a revision session be one topic.
The simplest version costs nothing: do past paper questions instead of topic sheets. An exam paper is already a mixed set, which is the whole point of it. Ours are free and sorted by board and year on the past papers pages, and if a whole paper is too much on a school night, five questions a day is a mixed set of five with the mark schemes beside them.
Two other things that help:
Get them to say out loud what kind of question it is before they start. That is the step being trained, and saying it makes it visible.
Judge a revision session by what is still there next week, not by how it felt at the time. A session that felt hard and left something behind beat a session that felt good and did not.
The short version
Working through one topic at a time makes practice feel productive because it removes the hardest decision in the exam, which is working out what the question is asking for. Mixing topics puts that decision back. It feels worse, it takes no extra time, and on a test weeks later it was worth more than twenty percentage points in both trials that measured it.
Sources: Rohrer, D., Dedrick, R. F., Hartwig, M. K., & Cheung, C.-N. (2020). A randomized controlled trial of interleaved mathematics practice. Journal of Educational Psychology, 112(1), 40 to 52. Rohrer, D., Dedrick, R. F., & Burgess, K. (2014). The benefit of interleaved mathematics practice is not limited to superficially similar kinds of problems. Psychonomic Bulletin & Review, 21(5), 1323 to 1330.