Blocked practice: the default, and the problem with it
Most studying happens in blocks: do 20 practice problems on one topic in a row, then move to the next topic and do 20 more on that, and so on. This is called blocked practice, and it's the intuitive default — it feels organized, and performance during the block itself tends to look strong, because working repeatedly on one type of problem lets a specific method or pattern get temporarily locked into working memory. The problem shows up later, on a test that mixes problem types together without labeling which method applies to which: the skill built during blocked practice was partly "which method to use," and blocking answers that question for free by only ever presenting one method at a time.
Interleaving: mixing instead of blocking
Interleaved practice mixes different topics or problem types together within the same study session, rather than grouping them into separate blocks — a few algebra problems, then a few geometry problems, then back to algebra, rather than finishing all the algebra first. This forces something blocked practice never requires: figuring out which method or concept applies to a given problem before applying it, because the type of problem isn't predictable from what came immediately before.
The counterintuitive research finding
Multiple controlled studies comparing interleaved and blocked practice — across math problem types, categorizing artwork by artistic style, and other skill domains — find a consistent pattern: blocked practice produces better performance during the practice session itself, but interleaved practice produces better performance on a later test, often by a substantial margin. Performance during practice and performance on a later, real test are not the same thing, and blocked practice's apparent advantage while studying doesn't carry over to when it actually matters.
This creates a genuine trap: interleaved practice often feels less effective while it's happening, because it's harder in the moment and progress feels slower or less smooth — but that felt difficulty is misleading feedback about what's actually being learned, similar to how active recall (see the companion article on that topic) feels harder than re-reading while working better in the long run.
Why interleaving works better
The leading explanation is that interleaving forces two skills at once instead of one: applying a specific method, and — separately — recognizing which method a given problem calls for. Blocked practice never has to build that second skill, because the block already announces which method applies before any individual problem is attempted. On a mixed real exam, that second skill (recognizing which approach a problem needs) is often the harder and more exam-relevant part, and blocked practice provides zero preparation for it, no matter how many repetitions were done within each block.
What interleaving looks like in practice
Interleaving doesn't require a complicated system — it mainly means resisting the pull toward grouping practice by topic:
- Mixing problem types within a single study session, instead of a solid block of one type followed by a solid block of another.
- Returning to earlier topics periodically rather than considering a topic "done" once its block is finished — this also connects interleaving to the spacing effect (see the companion article on that topic), since revisiting earlier material later is itself a form of spaced review.
- Using mixed-topic practice sets or past exams deliberately, rather than only topic-specific problem sets, in the run-up to a test that will itself be mixed.
When blocked practice still has a place
Interleaving isn't a strict universal replacement for blocked practice — when a method or concept is being encountered for the very first time, some initial blocked repetition to get the basic mechanics working is often necessary before interleaving that new material with others makes sense; mixing in something not yet minimally understood just produces confusion rather than the productive difficulty interleaving is meant to create. The general pattern that tends to work well: some blocked practice to establish a new skill, followed by interleaved practice once several skills are each individually functional, to build the discrimination between them that a real, mixed test will actually require.