The interleaving study technique: why mixing topics prepares you better for exams

The interleaving study technique means practicing with topics or problem types mixed together, instead of finishing one completely before moving on to the next. Rather than doing ten problems on topic A, ten on B and ten on C (that's called blocked practice), you shuffle them: one from A, one from C, one from B, another from A. It's harder and it feels worse. But on the exam, where nobody tells you which topic each question is from, it tends to pay off.

The reason is simple: blocked practice trains you to apply a method you already know you need. If you're working through the chapter 4 problem set, the problem is from chapter 4. The exam won't tell you what topic each question is, and half the difficulty is right there: recognizing what's being asked before you figure out how to answer it.

In this post I'll go over what the evidence says (it's good, but not for everything), how I use it, and an example from an intro management course, where mixing up cases from different theorists changes how ready you are on exam day.

What it is and why it works

Interleaving isn't studying lots of things at once with no plan. It's practicing in a way that forces you, on every question, to decide first what it's about. That decision is exactly what blocked practice spares you and exactly what the exam demands.

The best-known experiment is in math. In a study by Rohrer and Taylor, students who practiced mixed problem types did much better on the final test than students who practiced each type in its own block, even though they did worse during practice. That detail matters: blocked practice makes you feel like you've got it, because each problem goes faster than the last. But it goes faster because the topic was decided in advance.

It's not just numbers. In an experiment with paintings, people got better at recognizing each artist's style when the paintings were shuffled together than when they were grouped by artist. The interesting part: most participants believed they'd learned as well or better with the grouped paintings. Feeling harder doesn't mean it's working worse, and that's exactly why hardly anyone does it on their own.

When it works and when it doesn't

It's worth being honest here, because interleaving sometimes gets sold as a trick for everything. A review of 59 studies found that the effect is moderate and depends a lot on the material: it helps most when the things you're mixing are similar and easy to confuse, with complex tasks, and with visual material or problems. With expository texts the results were unclear, and with word lists, blocked study actually came out ahead.

In terms of what you're studying:

• It helps a lot for telling similar things apart: theorists and schools of thought that blur together, problem types that are solved differently, neighboring concepts (fixed cost vs. sunk cost, efficiency vs. effectiveness, inflation vs. devaluation).
• It helps with practice problems once you understand each topic on its own.
• It doesn't help yet when a topic is brand new to you. The first time, study the topic in a block until you get it; interleaving comes later, when you practice.
• It's not for memorizing a list of standalone definitions or reading a chapter for the first time.

An example from Principles of Management

In an intro management course, you're often asked to recognize which school or theorist an idea or a case belongs to: Taylor and scientific management, Fayol and the management process, Mayo and human relations, Weber and bureaucracy. If you study them in blocks, everything fits: you read the Taylor chapter and every example is Taylor. The exam, though, gives you a case and asks which approach it shows. I went into more detail in how to study Principles of Management.

Here's how I'd set it up:

1. First, one topic at a time. One card per theorist with the core idea, two or three key concepts and an example of your own. Up to here, blocked.
2. Then, mixed cases. Write (or pull from past papers) eight to ten short cases, two or three lines each, without saying which theorist they're about. For example: "A factory times every movement workers make and pays per piece produced." "A company finds productivity goes up when employees feel someone cares about them." "A government agency defines every position in writing, its duties and who it reports to."
3. Shuffle them and answer two things for each: which approach it is and which clue in the case told you. The clue is the part that teaches you the most.
4. Check and reshuffle. The ones you mixed up (Taylor vs. Fayol is the classic) go back in the pile.

If your course has problems instead of cases (accounting, statistics, corporate finance), it's the same idea: make a sheet with problems from three or four different topics, with no topic heading at the top, and before solving each one write down which method you'll use and why.

How to set up an interleaved session

Here's an approach that works for almost any theory-heavy or problem-based course:

• Pick three topics you've already covered that are easy to confuse. Three is plenty; with seven it turns chaotic.
• Prepare the questions in advance, mixed. They can be your own, from past papers or from the end of each chapter. What matters is that they're not grouped by topic.
• Answer without looking at your notes. Interleaving works best alongside active recall: pulling the answer from memory, not from the slide.
• When you get one wrong, write down what you confused with what. "I thought it was Fayol because it mentioned planning, but the stopwatch was the Taylor clue." That one line is gold before the exam.
• Repeat on another day, with the questions in a different order. Interleaving and spacing go well together: spaced repetition decides when you come back to a topic, and interleaving decides what you mix it with.

If you study from lecture slides, a quick way to collect the questions is the one I described in how to make flashcards from lecture slides: one card per idea, then shuffle them across lectures instead of reviewing one deck at a time.

Common mistakes

• Interleaving before you understand. If you don't yet know what each theorist says, mixing them just confuses you. Blocked study has its moment: at the start.
• Mixing courses instead of topics. Switching between accounting and history every half hour isn't interleaving: there's nothing to tell apart between them. What helps is mixing similar things within the same course.
• Giving up because it feels worse. It's normal for your first mixed session to be slower and full of mistakes. That's the effect, not a problem.
• Leaving the headings on. If the problem set says "Topic 3: depreciation" at the top, you're not interleaving anymore.

Doing it with your own lectures

Putting together mixed questions by hand takes time, and it's the reason almost nobody does it. In Filmina, a web app that explains your lecture slides out loud, I do it from Practice: I pick the Mixed type and, under What do you want to practice?, The whole subject or Choose several files, and the questions come from all those lectures together, like on an exam. What you miss comes back later. If you'd rather not use any app, a well-shuffled stack of index cards does the same job.

Frequently asked questions

What is interleaving in studying?

It's practicing similar topics or problem types mixed together in the same session, instead of finishing one before moving to the next. It makes you recognize what each question is about before answering it, just like on the exam.

Is it better to study one topic at a time or mix them?

To understand a new topic, one at a time. To practice what you already understand, mixed: the evidence shows it pays off more on the test, especially when the topics are similar and easy to confuse.

Does interleaving work for theory-heavy courses?

It works when you need to tell similar things apart, like theorists, schools of thought or neighboring concepts, and you practice with cases or questions. For reading a text for the first time or memorizing standalone definitions, the results are unclear.

How many topics should you interleave?

Three topics you've already covered and that are easy to confuse is enough to start. More than that makes the session chaotic and hard to check.