Most practice sessions are blocked: you do ten problems on topic A, then ten on topic B, then ten on topic C — AAAA BBBB CCCC. It feels organized, and you can usually solve each block of problems with increasing speed as you go. Interleaving mixes the topics together instead — ABC ABC ABC — and it feels noticeably harder, with more mistakes and a slower pace. The counterintuitive part is that interleaving produces better long-term retention and, more importantly, better performance on tests that mix topics the way real exams do.

What the research shows

In studies comparing the two approaches, students who practiced math problem types in blocks scored higher on practice during the session itself — they were, after all, doing the same type of problem repeatedly, so they got faster. But on a test given days later that mixed problem types together, the interleaved group substantially outperformed the blocked group, even though the blocked group had felt more confident during practice.

The gap exists because blocked practice teaches you how to solve a problem type when you already know which technique to use — the hard part has been removed. Interleaved practice teaches the part that’s actually hard on a real exam: recognizing which technique a problem calls for in the first place.

Why interleaving works: it trains discrimination, not just execution

Every exam — the SAT, IELTS, a final exam — presents questions in essentially random order with respect to topic. The skill being tested isn’t only “can you solve a system of equations” but “can you recognize that this problem is a system of equations problem when it’s sitting next to nine other problem types.” Blocked practice never trains that recognition skill, because the topic is given away by the structure of the practice session itself.

Interleaving also forces you to keep retrieving the underlying method for each topic repeatedly throughout a session rather than once at the start — which overlaps with the spacing effect described in spaced repetition, since each topic gets revisited at an interval rather than all at once.

How to interleave your own practice

You don’t need special materials — you need to deliberately mix what you already have:

  • For SAT Math, instead of doing 15 algebra problems, then 15 geometry problems, then 15 statistics problems, pull 5 of each and shuffle them into one set of 15. Work through them in the shuffled order. The SAT practice questions page is structured to support this.
  • For IELTS, instead of practicing only Reading passages in one sitting, alternate short blocks: one Reading passage, then a Writing Task 2 outline, then a set of Listening questions. This also mirrors the actual exam day better than single-skill marathons.
  • For any subject with a textbook, take the end-of-chapter problems from three different chapters you’ve already covered and combine them into a single mixed set before your next study session.

Common mistakes

Interleaving before you have basic fluency with each topic. If you don’t yet know how to solve a quadratic equation at all, mixing it with four other problem types just produces frustration without the discrimination benefit — you need a baseline level of competence in each topic first. Use blocked practice to build that baseline, then switch to interleaving to sharpen it.

Quitting because it feels like you’re doing worse. Interleaved practice produces more visible errors during practice — that’s the friction that builds the discrimination skill. Judge it by performance on delayed, mixed tests, not by how smooth the practice session felt.

How to start today

Take your next practice set and, instead of working through it in the order it’s printed, write down the topic of each problem, shuffle the list, and solve them in that new order. That’s the entire change — no new materials, no new schedule, just a different order.

FAQ

Should beginners use blocked practice first? Yes. Build basic competence in a topic with blocked practice, then shift to interleaving once you can solve problems of that type reliably on their own. Interleaving sharpens skills you already have more than it teaches brand-new ones.

Does this apply to language learning, like IELTS vocabulary or grammar? Yes — mixing grammar points or vocabulary topics in a single review session (rather than drilling one grammar rule for 20 minutes) better matches how language is actually used, where any rule might be needed in any sentence.

How does this relate to spaced repetition? They solve different problems and combine well: spacing controls when you revisit material over days and weeks, while interleaving controls how topics are ordered within a single session. A spaced-repetition queue that mixes due items from different topics is, in effect, already interleaved.