한국어

Interleaving Study Method: Does Mixing Topics Work?

At a glance
  • Interleaving means mixing different kinds of problems in one session. It worked well for math and for telling paintings apart.
  • College students who mixed problems scored 63% a week later; those who practiced one kind at a time scored 20%.
  • Mixing lowers your scores while you practice (60% against 89%). Feeling that it is going badly is normal.
  • The gain comes from choosing which method each problem needs, so it is largest when the kinds look alike.
  • For learning words, where there is nothing to choose, practicing in blocks did better.

Interleaving is a study method where you mix different kinds of problems in one session instead of finishing one kind before the next. For math and for telling categories apart, it worked well. In one experiment, college students who mixed their practice scored 63% on a test a week later; students who practiced one kind at a time scored 20%. The catch is that mixing feels worse while you do it, and it does not help with memorizing words.

What is the interleaving study method?

Interleaving means practicing different kinds of problems in a mixed order rather than in blocks of the same kind.

Most textbooks are built for blocked practice. After the lesson on the Pythagorean theorem come twelve problems that all use the Pythagorean theorem. Interleaving changes the order: instead of A, A, A, B, B, B, C, C, C you work A, C, B, C, A, B. The problems are the same, and so is the time. Only the order changes.

Why would order matter? In a block you know which method to use before you read the problem, because it is the same one as last time. An exam does not tell you. You first have to decide what kind of problem this is. Interleaving makes you practice that decision.

Does interleaving work?

Yes, for math it has been shown many times. In one experiment the gap a week later was more than threefold.

Rohrer and Taylor (2007) taught 18 college students how to find the volume of four unfamiliar solids. The students then worked the same practice problems in two sessions a week apart, either grouped by solid or mixed. The test came one week after the second session.

Blocked practice won during practice; mixed practice won the test a week later

  • Blocked by type
  • Mixed
0%25%50%75%100%Blocked by type · During practice · 89%89%Mixed · During practice · 60%60%During practiceBlocked by type · Test a week later · 20%20%Mixed · Test a week later · 63%63%Test a week later63% vs 20%
Mean accuracy of 18 college students on volume problems for four solids. Practice is the average of two sessions a week apart; the test came one week after the second session.Source: Rohrer & Taylor (2007), Experiment 2 results
See the numbers
Blocked by typeMixed
During practice89%60%
Test a week later20%63%

During practice the blocked group was far ahead, 89% against 60%. A week later it flipped: 63% for the mixed group, 20% for the blocked group. Eighteen students is a small sample, so read this alongside the classroom studies below rather than on its own.

Classrooms showed the same direction. Rohrer and colleagues (2015) gave 126 seventh graders the same math problems over three months, half in a mixed order and half in blocks. Everyone did the same review at the end, then took a surprise test either 1 or 30 days later.

Thirty days later, students who mixed scored 74% and students who blocked scored 42%

  • Blocked
  • Mixed
0%25%50%75%100%Blocked · Test after 1 day · 64%64%Mixed · Test after 1 day · 80%80%Test after 1 dayBlocked · Test after 30 days · 42%42%Mixed · Test after 30 days · 74%74%Test after 30 daysMixed wentfrom 80% to 74%
Mean accuracy of 126 seventh graders on a surprise test after three months of the same practice problems in different orders and the same final review.Source: Rohrer, Dedrick & Stershic (2015), results
See the numbers
BlockedMixed
Test after 1 day64%80%
Test after 30 days42%74%

Mixing was ahead after one day (80% against 64%), and further ahead after a month (74% against 42%). For the mixed group, the 1-day and 30-day scores differed by only 6 percentage points. Thirty times the delay, and they forgot almost nothing.

In a larger trial across 54 classes and 787 students who practiced for four months, the mixed group scored 61% on a test a month later and the blocked group 38% (Rohrer et al., 2020).

Why does interleaving feel harder?

Because every problem makes you choose a method, and that effort is where the benefit comes from.

Taylor and Rohrer (2010) had fourth graders practice four kinds of math problems, mixed or blocked, while keeping the gap before meeting the same kind again equal. The design asked whether mixing helps only because it spaces practice out. Mixing lowered practice scores but doubled scores on a test one day later. The errors showed why: the mixed group was better at pairing each problem with the right procedure.

Learners do not feel this. Kornell and Bjork (2008) showed people paintings by several artists, either one artist at a time or mixed with other artists. When asked who painted new paintings they had not seen, the mixed group did better. Yet participants rated blocked viewing as more effective, even after their own test results showed the opposite.

So when a mixed session feels like it went badly, that is not failure. The smooth feeling of blocked practice comes from having just used the same method. The same trap, where dropping the cards you got right also drops what you remember a week later, is in the note on whether flashcards are effective.

Does interleaving work for every subject?

No. It helps most when you must tell confusable things apart, and it hurt when there was nothing to tell apart, as with word lists.

Brunmair and Richter (2019) combined 59 studies with 158 samples in a meta-analysis. Overall, mixing had a moderate advantage (Hedges' g = 0.42). But it depended a lot on the material.

MaterialEffect size (g)Meaning
Paintings0.67mixing clearly better
Overall0.42mixing moderately better
Math problems0.34mixing a little better
Expository texts, tastesunclear (not statistically significant)no clear difference
Words-0.39blocking better

The analysis also found when mixing helps more: when categories look more alike, when items within a category differ more, and when the material is more complex. Put simply, mix when the hard part is "is this one or that one?" Learning English words with their meanings has no such choice, so mixing has little to give.

Dunlosky and colleagues (2013), reviewing ten study techniques, rated interleaved practice as moderate utility: promising enough to recommend, but not yet tested across as many conditions as practice testing and spaced practice.

How do you use interleaving, starting today?

Learn a new topic with a few blocked problems first, then mix it with problems from earlier topics. All you need is a textbook and a pencil.

  1. Start a new topic with a few problems in a row. If you do not know the method yet, you cannot practice choosing it. Every study above taught the methods before mixing.
  2. From the next session on, mix in older topics. Put a few of today's problems on one sheet with a few from last week's and last month's chapters. Sticky notes on several textbook pages work too.
  3. Name the kind of problem before you solve it. Say "this is a volume problem, not an area problem," pick the method, then solve. This one step is where interleaving does its work.
  4. Pair the things you confuse. Permutations and combinations, speed and acceleration: put the pairs you mix up next to each other.
  5. Do not bother mixing word lists. Spend that time recalling before you check, and recalling again a few days later. Review gaps are covered in when to review with spaced repetition.

In Aloom

Aloom's review does not come in topic blocks. When the answers you keep from asking AI become nodes in your knowledge graph, each day brings the number you chose, 1, 3 or 5, picked by when each one is due rather than by topic. So today's review mixes answers kept on different days about different things, and each question starts with "which of my topics was this?" The more you recall a node, the darker its ink. Take a look at Aloom

FAQ

What is the difference between interleaving and blocked practice?

Blocked practice works through one kind of problem before moving on (AAA BBB CCC). Interleaving mixes the kinds (ABC BCA CAB). Blocked practice scores higher during practice; interleaving scored higher on later tests in the math studies above.

Is interleaving the same as spaced practice?

No. Spacing returns to the same material after a gap; interleaving mixes different material. Mixing spreads out each kind on its own, so the two often go together, but Taylor and Rohrer (2010) kept the spacing equal and still found mixed practice doubled test scores.

What subjects is interleaving good for?

Subjects where you must choose among methods or tell categories apart: math and physics problems, or sorting paintings and images by kind. For word learning, blocking did better.

References