Interleaved practice: why mixing subjects beats studying one thing at a time

Finishing one subject before starting another feels orderly: algebra on Monday, grammar on Tuesday, biology on Wednesday. Interleaved practice instead mixes problem types when the task requires choosing which method to use. Practice order can affect what is remembered later.

Interleaved practice mixes problem types or topics within a study session. It needs no special tool, but it is not always the best choice: the content and the learner’s familiarity matter. Below we examine the evidence, its limits, and a way to apply the idea with questions from different sources.

The illusion of blocked study

In blocked study, you repeat the same type of exercise until you feel you have mastered it: twenty ratio problems, then twenty percentage problems, then twenty on compound interest. Fluency builds quickly, errors drop as the list goes on, and the sense of progress is immediate. The problem is that much of that fluency comes from context, not knowledge. If every question on the list is a ratio problem, you never have to decide which method to use: the chapter title already made that decision for you.

On an exam, no one tells you which formula a question calls for. And that step, looking at an unfamiliar problem and identifying what kind it is, is exactly what blocked study never trains. Interleaving trains precisely that: by alternating fractions, equations, and geometry within the same list, you are forced to discriminate each problem before solving it. That extra step is uncomfortable, and it is what produces the kind of learning that survives until exam day.

What the studies show (including where they surprise you)

The experiment that influenced me most was Rohrer and Taylor (2007). College students learned to calculate the volume of four geometric solids. Half practiced in blocks, one solid at a time; the other half practiced with shuffled problem sets. During practice, the shuffled group made far more errors and almost certainly walked away feeling they had learned less. A week later, on the final test, the picture flipped dramatically: the students who had practiced in mixed order performed far better. Stumbling during practice was the price of succeeding when it counted.

That result did not stay in the lab. In a later study run in real mathematics classrooms, Rohrer, Dedrick, and Stershic (2015) assigned interleaved or blocked worksheets across months of middle-school lessons. The interleaving advantage showed up on the tests and was largest precisely on the assessment furthest from practice, the scenario that most resembles a real exam, where the material has long gone cold.

And the effect is not limited to math. In the classic experiment by Kornell and Bjork (2008), participants studied paintings in order to learn to recognize each artist's style. Seeing the works interleaved, rather than grouped by painter, made participants better at classifying new paintings they had never seen. That ability to generalize to novel examples is exactly what a well-designed exam demands.

The detail almost nobody mentions: your intuition will vote against it

To me, the most uncomfortable finding in the Kornell and Bjork study was not the performance data, it was the judgment data. After the test, with interleaving's superiority already demonstrated in their own results, most participants still claimed that blocked study had worked better for them. The technique wins, and the people who used it believe it lost.

Ease during a session is not a reliable measure of retention. Robert Bjork uses the term desirable difficulties for conditions that can make immediate practice harder while benefiting later tests. When trying interleaving, compare performance after a delay; a lower score right away does not by itself prove better learning.

Where interleaving fails (and when not to use it)

Here is the part that generic articles on this topic almost never include: interleaving has well-defined boundary conditions.

• Absolute beginners: if you cannot yet solve a basic problem of a given type without checking your notes, mixing will only produce confusion. Start with a short, focused block; as soon as you get your first unaided solution, switch to mixing.
• Unrelated content: the main gain of interleaving comes from learning to discriminate between similar, confusable types. Mixing topics nobody would ever confuse, such as Brazilian history and English verb conjugation, still gives you the benefit of spacing, but loses the discrimination effect that is the heart of the technique.
• Confusing interleaving with multitasking: switching subjects every two minutes with notifications in between is not interleaving, it is fragmentation. Each round needs enough time for genuine effort; fifteen to thirty minutes per topic is an illustrative block, not a proven ideal for everyone.
• The evidence has known limits: in the major review of study techniques by Dunlosky and colleagues (2013), interleaved practice received only a moderate utility rating, not because it failed in experiments, but because at the time the evidence base was concentrated in mathematics and category learning. The honest claim is that the technique is highly promising and especially strong in those areas, not that it is foolproof for every kind of content.

Interleaving, retrieval practice, and spacing are the same engine

Interleaving, retrieval, and spacing are related processes, but they are not interchangeable. Returning to a topic after another creates an interval; answering without the source involves retrieval. Dunlosky’s review rated practice testing and distributed practice highly. The specific benefit of mixing topics still needs to be judged by task type.

How to practice interleaving with SimulAI Academy

A test taken immediately after reading a single chapter may reflect recent familiarity. To check retention, also answer questions after a delay and mix topics studied earlier. Compare errors by topic rather than treating one session’s score as proof of lasting learning. One possible workflow is:

• Generate quizzes from more than one source: upload the PDFs or notes from two or three topics you have already studied and build a mixed mock exam. A question that arrives without a chapter label forces the discrimination step the studies describe.
• Mix old and new flashcard decks: before reviewing this week's deck, interleave cards from previous weeks. Stumbling on an old card is worth more than ten easy hits on cards you created an hour ago.
• Use the mind map to choose what to mix: the best candidates for interleaving are neighboring concepts you confuse with each other. When two branches of the map feel interchangeable in your head, that is the signal they need to appear side by side in your questions.
• Let the material go cold: a gap of a few hours, or a day, between reading something and taking a quiz on it turns the test from a reading-comprehension check into genuine retrieval training.

Shuffling questions is technically simple; deciding when to do it takes care. If basic concepts are still unclear, first practice examples of each type. Then mix types and see whether the learner can choose the right procedure. An initial drop in score does not automatically mean that the method worked or failed.

A starter plan for this week

• Pick two or three topics you have studied at least once and that are somewhat related to each other.
• Build, or generate, a mixed question list with no grouping by subject.
• Before solving each question, jot down in one word what type it is. That small ritual trains exactly the discrimination the exam will demand.
• Expect to make more errors than in blocked study, and do not abandon the technique because of it: the studies show this is the expected phase of the process.
• Redo a mixed list on the same topics a few days later and compare the result with your first attempt, not with how the blocked session felt.

Mixing topics can make practice feel less fluent. Use a later review to decide whether interleaving helped with the material studied. Be wary both of immediate ease and of claims that one strategy fits every subject.

References

• Rohrer, D., & Taylor, K. (2007). The shuffling of mathematics problems improves learning. Instructional Science, 35, 481–498. https://link.springer.com/article/10.1007/s11251-007-9015-8
• Rohrer, D., Dedrick, R. F., & Stershic, S. (2015). Interleaved practice improves mathematics learning. Journal of Educational Psychology, 107(3), 900–908. https://psycnet.apa.org/doi/10.1037/edu0000001
• Kornell, N., & Bjork, R. A. (2008). Learning concepts and categories: Is spacing the "enemy of induction"? Psychological Science, 19(6), 585–592. https://journals.sagepub.com/doi/10.1111/j.1467-9280.2008.02127.x
• Dunlosky, J., Rawson, K. A., Marsh, E. J., Nathan, M. J., & Willingham, D. T. (2013). Improving Students' Learning With Effective Learning Techniques. Psychological Science in the Public Interest, 14(1), 4–58. https://journals.sagepub.com/doi/10.1177/1529100612453266