Interleaving: Why Mixed Practice Beats Blocked

What Is Interleaving?

Interleaving is a study strategy where you mix different types of problems or topics in a single practice session instead of grouping them into separate blocks. If you're learning to calculate the volume of solids, blocked practice means solving ten problems on cubes, then ten on cylinders, then ten on cones. Interleaved practice means shuffling all thirty so a cube problem is followed by a cone, then a cylinder, in no fixed order.

That small change in ordering has an outsized effect on how much you remember later. The reason is subtle. When problems are blocked, you already know which method to use before you read the question, because the last five problems used the same one. You're just repeating a procedure. When problems are interleaved, every question forces a decision first: what kind of problem is this, and which approach does it need? That decision is the part that transfers to a real exam, where problems never arrive pre-sorted by topic.

Interleaving belongs to a small family of evidence-based techniques that share a strange feature: they make practice feel harder and less productive while making learning more durable. Robert Bjork groups them under the label desirable difficulties, conditions that slow you down now and pay off later. Interleaving is one of the best-documented members of that family, and also one of the most counterintuitive.

The Evidence: Six Studies, One Pattern

The research on interleaving is unusually clean, because the same result keeps showing up across labs, ages, and materials. Here's the core evidence, from the first controlled study to a large randomized trial in real classrooms.

Rohrer and Taylor kicked it off in 2007 with "The shuffling of mathematics problems improves learning" (Instructional Science, 35(6), 481-498). College students practiced volume formulas in blocked or interleaved order. During practice, the blocked group looked far better, solving nearly every problem correctly. On a test a week later, that advantage flipped hard: the interleavers dramatically outperformed the blockers. Practice performance and real learning pointed in opposite directions.

Then came the classroom studies. Taylor and Rohrer (2010, Applied Cognitive Psychology, 24, 837-848) had fourth graders practice four kinds of problems. One day later, the interleaved group scored 77% versus 38% for the blocked group, an effect size of d = 1.21. That's not a nudge; interleaving roughly doubled the score.

Rohrer, Dedrick, and Stershic (2015, Journal of Educational Psychology, 107(3), 900-908) ran 126 seventh graders over three months. The interleaved advantage on a delayed test reached d = 0.79, and interleaving showed something remarkable: near-immunity to forgetting. When the delay before testing grew 30-fold, blocked scores collapsed while interleaved scores fell only from 80% to 74%.

The capstone is Rohrer, Dedrick, Hartwig, and Cheung (2020, Journal of Educational Psychology, 112(1), 40-52), a randomized controlled trial with 787 students across 54 classes in five schools. Students practiced interleaved or blocked assignments over four months, then took an unannounced test one month later. Interleaved beat blocked, 61% versus 38%, with a large effect size of d = 0.83 (95% CI [0.68, 0.97]).

Zoom out and the picture holds. Brunmair and Richter's 2019 meta-analysis, "Similarity matters" (Psychological Bulletin, 145(11)), pooled 59 studies, 238 effect sizes, and 158 samples. The overall interleaving benefit was moderate and reliable, g = 0.42. Dunlosky and colleagues' influential 2013 review rated interleaved practice as a technique of "moderate utility," held back only because it had fewer classroom trials at the time. The trials since have made the case stronger, not weaker.

Why Interleaving Actually Works

Two mechanisms explain why shuffling helps, and they aren't mutually exclusive.

The first is discrimination, sometimes called the discriminative-contrast hypothesis. When you see a cube problem right after a cone problem, the contrast makes the defining features of each type pop. You learn not just how to solve a cube problem but how to recognize one. Kornell and Bjork (2008) argued that mixing examples exposes the differences between categories, and Birnbaum, Kornell, and the Bjorks (2013) put it directly: interleaving helps because doing so "highlights differences between the categories." Blocking hides those contrasts, because everything on the page is already the same kind.

The second is retrieval. In a blocked set, you load a strategy into working memory once and reuse it. In an interleaved set, you have to retrieve the right approach from long-term memory for every single problem, because the previous one won't tell you. That repeated act of retrieval, choosing and pulling up a method under mild uncertainty, is itself a powerful learning event. It's the same engine behind active recall and the blurting method: the effort of retrieval, not the exposure to information, is what builds durable memory.

Put those together and interleaving trains the exact skill that blocked practice skips. Real problems, whether on an exam, in a clinic, or on the job, don't come labeled. The hard part isn't running a procedure; it's knowing which procedure the situation calls for. Interleaving rehearses that judgment every few seconds. Blocking removes it entirely.

Interleaving vs. Blocking vs. Spaced Repetition

Interleaving gets confused with spacing constantly, partly because both feel harder than the default and both show up on the same lists of desirable difficulties. They're related but distinct.

Blocking is the default: study one thing until you've got it, then move on. It maximizes how good you feel and how fast you improve within the session, and it's fine for the very first exposure to a brand-new skill.

Interleaving is about the order of practice within a session: you mix related but different topics so consecutive items require different responses. It targets your ability to discriminate and select.

Spaced repetition is about the timing of review across sessions: you revisit material after growing intervals so you retrieve it just as it's starting to fade. It targets long-term retention of individual items. If you want the full mechanics, see spaced repetition for readers.

The practical takeaway: these stack. A strong routine uses blocking for the very first exposure, interleaving once you have the basics of several related topics, and spacing to carry it all across weeks. They solve different problems, so you don't have to choose one.

The Metacognitive Trap

Here's the reason interleaving stays rare despite decades of evidence: it feels like it isn't working, and your feelings are convincing.

Kornell and Bjork's 2008 study "Learning Concepts and Categories: Is Spacing the 'Enemy of Induction'?" (Psychological Science, 19(6)) is the sharpest demonstration. Participants studied paintings by different artists, either massed (all of one artist's works together) or interleaved (mixed among other artists). Then they saw new paintings and had to name the artist. Interleaving won clearly; people were better at identifying the style of a painter they'd studied in mixed order.

The twist is what people believed. When asked which method helped them more, learners overwhelmingly picked massing, the one that had actually hurt them. Of the participants who expressed a preference, roughly nine in ten (64 of 72) said the blocked schedule was more effective, even right after their own test scores proved the opposite. Their sense of learning was inversely related to their actual learning.

This is the fluency illusion at work. Blocked practice feels smooth because repetition makes each problem easier than the last, and your brain reads that smoothness as mastery. Interleaved practice feels bumpy because every problem is a small fresh struggle, and your brain reads the struggle as failure. Both readings are wrong. As the research on desirable difficulties shows again and again, the study experience that feels productive and the study experience that is productive are often opposites. If a technique feels a little uncomfortable and slow, that's frequently a sign it's doing something.

Where Interleaving Helps and Where It Backfires

Interleaving isn't magic dust you sprinkle on everything. The Brunmair and Richter meta-analysis found the effect depends heavily on what you're learning, and one condition is where the whole thing quietly matters most: similarity.

Interleaving pays off when the things you're mixing are similar enough to be confused. That's when discrimination has something to do. Their data showed the biggest gains for inductive category learning like paintings (g = 0.67) and a solid, reliable benefit for mathematics (g = 0.34). The effect shrank or vanished for materials with no confusable neighbors, and for pure word-list memorizing, blocking actually did slightly better (g = -0.39). If categories are wildly different, there's nothing to discriminate, so mixing them just adds switching cost with no upside.

Two guardrails follow. First, don't interleave before you have the fundamentals. A student who can't yet solve a single cube problem gains nothing from mixing cubes with cones; they need one clean block first, then mixing. Interleaving is for consolidation and discrimination, not first contact. Second, mix things that live in the same neighborhood. Interleaving Spanish verb conjugations with each other is smart. Interleaving Spanish verbs with calculus derivatives is just task-switching, and task-switching has real costs.

How to Interleave Your Own Studying

The technique is simple to apply once you know its shape. A few concrete moves:

• Shuffle problem sets by hand. Most textbooks block problems by section. Photocopy or list problems from the last three or four sections, then reorder them so no two in a row come from the same section. Practice from the shuffled set. This single habit captures most of the benefit for math, physics, chemistry, and statistics.
• Group by "what skill does this need," not "what chapter is this from." When you build a study set, deliberately include items that look similar but require different responses. The confusable pairs are where interleaving earns its keep.
• Rotate topics within a session. Studying for a biology exam covering three systems? Do a few questions on each, cycle through, and repeat, rather than finishing one system before starting the next.
• Interleave across subjects at a coarser grain. Alternating 40-minute blocks of math, then history, then Spanish over an afternoon gives you spacing plus mild interleaving without the cost of switching every 30 seconds. Keep the mixing tight within a subject and loose across subjects.
• Expect it to feel worse, and keep a scoreboard. Because interleaving lowers your in-session accuracy, you can't trust the "this feels productive" signal. Track delayed performance instead: quiz yourself a day or two later. The delayed score is the only honest measure.
• Don't mix apples and orbital mechanics. Keep interleaved items within a similar domain so there's something to discriminate. Random chaos isn't interleaving.

For anyone building a serious study system, interleaving pairs naturally with retrieval and spacing. Use interleaving to organize a session, use active recall as the mode of practice within it, and use spaced repetition to schedule the sessions. The combination is close to a complete evidence-based routine.

Interleaving for Readers and Lifelong Learners

Interleaving isn't only for students grinding through problem sets. If you read to learn, whether articles, books, or research, the same logic applies to how you review and connect ideas.

The trap for self-directed readers is the reading-world equivalent of blocking: you finish a book, feel the smooth glow of familiarity, and move on without ever retrieving or contrasting what you read. Highlighting can make this worse if it's passive, as the research in the science of highlighting shows. The fix is to turn your reading into interleaved, retrieval-based review, and that's where a highlighting and knowledge tool earns its place.

Here's a workflow that builds interleaving into how you read:

• Highlight across sources, then review them mixed. With Glasp's web highlighter, your highlights from many different articles and books collect in one place. Reviewing that mixed stream, rather than re-reading one source top to bottom, is interleaving by default: you're constantly switching contexts and retrieving what each highlight meant.
• Contrast similar ideas from different authors. Interleaving thrives on discrimination. When you save highlights on the same theme from competing sources, put them side by side and force the distinction: how does this author's claim differ from that one's? That's the discriminative contrast that makes interleaving work.
• Mix media. Alternate a written article with a YouTube video summary on a related topic. Switching format keeps retrieval effortful and prevents the fluency illusion that sets in when everything looks the same.
• Quiz your own notes out of order. Use Glasp's AI chat to ask questions across all your highlights, not one document at a time. Answering questions that jump between sources is retrieval practice and interleaving at once.
• Fold in your Kindle library. Pull your book highlights in with Kindle highlights so your review set spans articles, videos, and books together. A mixed pool is a better interleaved pool.

The deeper point connects to how to remember what you read: passive exposure, no matter how much you highlight, doesn't build durable memory. Mixing sources and retrieving across them does. Interleaving gives you a concrete way to structure that review instead of hoping it happens.

Frequently Asked Questions

Is interleaving the same as multitasking?

No, and the difference matters. Multitasking means splitting attention across tasks at the same time, which reliably hurts performance. Interleaving means giving full attention to one problem, finishing it, then switching to a different type of problem. You're sequential, not simultaneous. The switching is the point, but you're never doing two things at once.

Won't mixing topics just confuse me?

At first, yes, and that's expected. Interleaving lowers your accuracy and confidence during practice, which is why it feels confusing. But that confusion is the brain doing the work of telling problem types apart. The studies consistently show that the group who felt more confused during practice scored higher on delayed tests. The one real caveat: don't interleave before you've grasped the basics of each topic. Learn the fundamentals in a short block first, then mix.

How is interleaving different from spaced repetition?

Spaced repetition controls the timing of review: you space sessions out over days or weeks so you retrieve material just before forgetting it. Interleaving controls the order of practice within a session: you mix different problem types instead of blocking them. They're complementary. A good routine spaces sessions apart and interleaves topics inside each one.

Does interleaving work outside of math?

Yes, whenever the material has confusable categories. It's well documented for math and for inductive learning like recognizing painting styles, bird species, or medical images. It helps with grammar rules, vocabulary in a foreign language, and diagnostic reasoning. It helps least when topics are so different there's nothing to confuse, and it can slightly hurt for memorizing isolated word lists.

How much does interleaving actually improve scores?

In controlled classroom studies, a lot. Fourth graders scored 77% versus 38% (Taylor & Rohrer, 2010), and a large randomized trial found 61% versus 38% on a surprise test a month later (Rohrer et al., 2020). Across 59 studies, the average effect was moderate and reliable (Brunmair & Richter, 2019). Your mileage depends on the material, but the direction is remarkably consistent.

The Bottom Line

Interleaving is one of the most reliable findings in learning science, and one of the most ignored, because it commits the unforgivable sin of feeling unproductive. Blocking gives you the warm sensation of progress and leaves you with little a week later. Interleaving gives you a bumpy, uncertain practice session and durable skill you can actually use when problems arrive unsorted.

The move is small. Shuffle your practice so no two consecutive problems use the same approach, mix similar topics that are easy to confuse, and judge the method by your delayed scores rather than how it feels in the moment. Do that, and you're training the skill that matters: not running a procedure you were handed, but choosing the right one yourself.

If you learn mostly by reading, the same principle reshapes how you review. Instead of re-reading one source until it feels familiar, collect highlights across many sources with Glasp, mix them, contrast them, and quiz yourself across the whole pool. Familiarity is a feeling. Learning is what survives the mixing.