Interleaving: mix up what you practise#

Imagine you are learning to tell three painters apart—a Monet, a Cezanne, a Renoir. You could study all the Monets, then all the Cezannes, then all the Renoirs. It feels orderly, and while you are looking at the Monets you really do feel like you “get” Monet. Now you sit the test, paintings come at you in random order, and the feeling evaporates. The trick was never recognising one painter in isolation—it was telling them apart. Studying them in separate blocks gives you almost no practice at the thing the test actually asks for.

Interleaving is the fix: instead of practising one thing to exhaustion before moving on (AAABBBCCC—“blocking”), you mix related items together (ABCABC). You shuffle the deck. It feels messier and slower, you make more mistakes along the way, and that is precisely the point.

Blocking versus interleaving#

Most of us study in blocks by default. Do twenty quadratic-equation problems in a row, then twenty about volumes, then twenty about probability. Within a block you get into a groove—and the groove is the problem. After the first question you no longer have to work out which kind of problem you are facing; the textbook already told you. You just repeat the same procedure twenty times.

Interleaving breaks the groove. When the next problem could be any type, you have to do the harder, more honest piece of thinking first: what kind of problem is this, and which method does it call for? That is the skill you actually need when problems do not come pre-sorted—in the exam, in the cockpit, on the job.

The classic demonstration comes from maths. When learners practised mixed sets of problems rather than blocked sets, their performance during practice dipped, but their scores on a later test rose sharply (Rohrer & Taylor, 2007). Same problems, same time on task—only the order changed, and the order was worth more than the effort.

This is a close cousin of the contextual interference effect you may have met in skill learning—see Variation . Practising a motor skill under varied, jumbled conditions hurts you in the session and helps you afterwards, for the same underlying reason: easy, predictable practice lets you coast, and coasting does not stick.

Why it works: discrimination and retrieval#

Two things are happening when you interleave.

You learn to discriminate. Mixing similar categories side by side forces you to notice what makes each one distinct—the brushstroke, the telltale phrase in the problem, the feature that says “this is a that, not a this.” Blocking hides those contrasts because you never see the cases together. This is why interleaving helps most when the things you are mixing are confusable: bird species, tax rules, chord shapes, problem types that look alike but need different methods.
You practise retrieval and selection. Every time the topic switches, you have to pull the right approach back out of memory from cold, rather than keeping it warm and ready. That effortful reload is doing real work—it is the same mechanism that makes self-testing beat re-reading.

Notice the honest catch in both: interleaving makes practice feel worse. You will be slower, you will fumble, and your in-session score will be lower than if you had blocked. That sinking “I’m not getting this” feeling is not failure—it is the difficulty doing its job. Judge interleaving by a delayed test a few days later, never by how smooth the practice session felt.

Where it helps—and where it doesn’t#

Interleaving is powerful but it is not a universal switch to flip on everything. A large review of the research found the benefit is real but depends heavily on what you are mixing (Brunmair & Richter, 2019).

It shines when items are similar and confusable. Categories or problem types you keep mixing up are the sweet spot—that is exactly the discrimination interleaving trains.
It fades when items are unrelated. Shuffling French vocabulary with chemistry equations buys you little; there is nothing to discriminate between, so the extra effort is wasted.
Mind the cost of switching. If the task carries a heavy setup or “spin-up” cost each time you change to it, constant restarting can swamp the benefit. Interleave at a grain where switching is cheap.
Get a foothold first. When something is brand new, a little blocking to grasp the basic procedure before you start mixing is sensible—you cannot tell two things apart until you have a rough sense of each.

A practical recipe: learn the move in a short block, then interleave it with the neighbours it is easiest to confuse it with, in a randomised order, and space those sessions out over days. Interleaving, spacing and retrieval practice are siblings—they stack.

The takeaway#

Stop practising one thing to death before moving on. Shuffle related, easily-confused topics together so that every item forces you to ask “what is this, and what does it need?” before you answer. It will feel harder and your practice scores will dip—welcome that. The mess is what teaches your brain to tell things apart and to pull the right method out under pressure, which is the whole game when the questions stop arriving in neat, labelled blocks.