Interleaving Practice

Most students revise one topic at a time: complete all of Chapter 3, then move to Chapter 4, then Chapter 5. This is blocked practice — all problems of one type are finished before moving to the next.

Interleaved practice mixes topics within a single session. Instead of completing all quadratic equation questions before touching trigonometry, you alternate: one quadratic question, one trigonometry question, one probability question, another quadratic — and so on.

Blocked session (quadratic equations only):

Q	Question type	Topic
1	Expand (x + 3)(x − 2)	Quadratics
2	Expand (2x + 1)(x + 5)	Quadratics
3	Factorise x² + 5x + 6	Quadratics
4	Solve x² − 9 = 0	Quadratics

Interleaved session (mixed topics):

Q	Question type	Topic
1	Expand (x + 3)(x − 2)	Quadratics
2	Find angle θ in a right triangle with sides 5, 12, 13	Trigonometry
3	Factorise x² + 5x + 6	Quadratics
4	A bag contains 3 red, 2 blue. P(red, then blue without replacement)?	Probability

Interleaved practice is reliably harder and more uncomfortable than blocked practice. Students consistently underestimate how much they've learned from an interleaved session because the difficulty feels like poor performance rather than progress.

In a blocked practice session, you never have to decide which method applies to the current question. You already know — it's the method you've been practising for the past 30 minutes.

In an exam, that context is gone. Every question requires you to identify the relevant concept, formula, or technique before applying it. This is the discrimination problem — and interleaving trains it directly.

When topics are mixed, each question requires two cognitive steps:

1. Identify which strategy, formula, or concept is relevant
2. Apply it correctly

Blocked practice trains step 2. Interleaved practice trains both.

This explains a well-documented result: students who practise in interleaved blocks perform significantly worse during practice, but perform significantly better on delayed tests. Their performance during the session is a poor indicator of what they've actually retained.

Interleaved practice feels harder because it is doing more work. The difficulty is the learning — not an obstacle to it.

The underlying mechanism is called the contextual interference effect, first documented by Shea and Morgan (1979) in motor learning research, then replicated extensively in academic subject domains. The "interference" caused by switching between problem types forces the brain to reconstruct the relevant procedure each time, rather than executing a recently active routine from short-term memory.

A representative study by Taylor and Rohrer (2010) tested Year 7 students on four types of maths problems. One group practised in blocked sessions; the other practised with interleaved problem sets containing all four types mixed randomly.

During practice: the blocked group performed more accurately. Having just completed several problems of the same type, they were fluent with that method in the moment.

One week later: the interleaved group scored substantially higher than the blocked group on a test covering all four problem types.

The blocked group's higher accuracy during practice predicted almost nothing about their long-term retention. The interleaved group's lower accuracy during practice masked substantially stronger encoding.

If you want to judge how effective a revision session was, do not use how easily you answered the questions today. Use how well you answer questions on the same material next week.

This finding generalises beyond maths. The contextual interference effect has been replicated in foreign language vocabulary, science problem-solving, and motor skill learning. The principle is consistent: interleaving produces worse immediate performance and better delayed retention.