Interleaved Practice: Why Mixing Problem Types Works

Most practice sets are built in blocks. After a lesson on one skill, you get a page of problems that all use that skill. It feels efficient, and you usually get most of them right. A series of classroom studies led by psychologist Doug Rohrer tested a different layout, called interleaved practice, where problems of different kinds are mixed together. The results were large enough to be worth changing how you do homework.

Blocked vs interleaved, in plain terms

Blocked practice: ten problems in a row that all use the same method, such as ten Pythagorean theorem problems right after the Pythagorean theorem lesson.

Interleaved practice: the same kinds of problems, but shuffled so that consecutive problems call for different methods. One problem needs slope, the next needs a graph, the next needs a different formula.

The 2015 paper by Rohrer, Dedrick, and Stershic points out the hidden weakness of blocks: when every problem relates to the lesson you just finished, you know which strategy to use before you even read the problem. Interleaving removes that clue. You have to look at each problem and decide what kind it is, which is exactly what happens on a cumulative exam or in a later course.

What the 2015 classroom study found

In that experiment, 126 seventh-grade students worked the same practice problems over a three-month period. The only difference was the arrangement: some skills were practiced in the usual blocks, others in an interleaved order. The practice phase ended with a review session, and then students took an unannounced test either 1 day or 30 days later.

Interleaved practice produced higher scores on both tests. The effect sizes reported were 0.42 for the test one day later and 0.79 for the test 30 days later. The gap was larger after a month, which suggests interleaving matters most for what you still know when the exam finally arrives.

What a larger trial found

A 2020 study by Rohrer, Dedrick, Hartwig, and Cheung tested the idea at a much larger scale. It was a preregistered, cluster randomized controlled trial with 54 seventh-grade math classes. Over four months, classes periodically completed either interleaved or blocked assignments, and then both groups completed the same interleaved review assignment.

One month later, students took an unannounced test. The interleaved group scored 61 percent and the blocked group scored 38 percent, a difference of 23 percentage points, with an effect size of 0.83. Teachers ran the intervention without special training, and in an anonymous survey completed before they knew the results, they expressed support for interleaved practice. The authors still note that important caveats remain, so this is strong evidence, not a final word.

Why your textbook probably does not do this

The same research group counted how often textbooks actually mix problems. In a 2020 paper, they examined 13,505 practice problems in six representative mathematics texts and found that only 9.7 percent were interleaved. That works out to about one or two interleaved problems per school day.

So if you only do the assigned blocked sets, you are likely getting very little of the kind of practice these trials found most effective. The fix is something you can do yourself.

Why it feels worse while you do it

Interleaved practice is harder in the moment. You will get more problems wrong during practice, because you are now making a decision you used to skip. That can make blocked practice feel more productive. The trials above measured learning on later, unannounced tests, which is where interleaving pulled ahead. A rough practice session is not a sign the method is failing.

How to interleave your own homework

1. Finish the assigned problems first if your teacher grades them as given.
2. Build a mixed review set. Pull two or three problems from each of the last several sections or chapters, and shuffle them so no two neighbors use the same method.
3. Name the problem type before solving. Write a one-word label such as "slope," "system," or "percent" next to each problem. Choosing the strategy is the skill interleaving trains.
4. Check answers after the whole set, not after each problem, so you cannot use the answer key to guess which method a section wanted.
5. Keep mixing older material in. A problem type from a month ago belongs in today's set too. That also spaces your review of each skill across many days.

This works best for subjects where different problems look similar but need different methods, such as math, physics, and chemistry. The studies here were all in mathematics, so treat other subjects as a reasonable extension rather than a tested result.

How this differs from mixing examples with practice

You may have seen advice to alternate between reading a worked example and solving a problem yourself. That is a different idea. It mixes how you practice (studying a solution versus producing one). Interleaving, as tested in these studies, mixes what you practice: different kinds of problems in the same set.

Key takeaways

• Interleaved practice mixes different kinds of problems so consecutive problems need different strategies.
• In a 2015 classroom study of 126 seventh graders, interleaving beat blocked practice, with a larger advantage after 30 days.
• A 2020 trial with 54 classes found interleaved students scored 61 percent versus 38 percent on a surprise test a month later.
• Only 9.7 percent of problems in six representative math textbooks were interleaved, so you usually have to build mixed sets yourself.
• Label each problem's type before solving, and expect practice to feel harder even as later test results improve.