Blocked and interleaved
Blocked practice is AAA BBB CCC: all the problems of one kind, then all of the next. Most textbooks are built this way. Each chapter ends with problems on that chapter.
Interleaved practice is ABC BCA CAB: the kinds are mixed, so each problem is of a different kind than the one before it.
The difference is what you must do before you solve. In a block, you already know which method to use, because it is the same as last time. In a mix, you must first decide what kind of problem this is. A real test, and real life, ask that question every time.
What the studies found
| Year | Study | What it found |
|---|---|---|
| 2007 | Rohrer and Taylor, Instructional Science | Students practiced four kinds of geometry problem, blocked or mixed. During practice, the blocked group scored higher. On a test one week later, the mixed group scored 63 percent and the blocked group 20 percent. |
| 2008 | Kornell and Bjork, Psychological Science | People learned the styles of twelve painters. Paintings shown mixed led to better recognition of new paintings than paintings shown one painter at a time. Most participants still believed the blocked order had taught them more. |
| 2015 | Rohrer, Dedrick and Stershic, Journal of Educational Psychology | In seventh-grade classrooms, mixed math practice gave about 74 percent on a test 30 days later. Blocked practice gave about 42 percent. |
| 2019 | Brunmair and Richter, Psychological Bulletin | A meta-analysis. The benefit is moderate overall. It is largest for visual categories such as paintings, positive for math, and negative for lists of words. |
Why it works
Two reasons have support. The first is discrimination: when two kinds sit side by side, you see what makes them different. That is the title of the 2019 review, “Similarity matters”: the effect is strongest when the kinds are easy to confuse.
The second is spacing and retrieval. In a mix, each return to a kind comes after a gap, so you must pull the method out of memory again. That is retrieval practice inside the practice.
Where it does not work
- When the material is brand new. You need a first understanding of each kind before a mix helps. Learn one, then mix.
- When the kinds have nothing in common. Mixing French verbs with chemistry does not teach you to tell anything apart. Mix things that can be confused.
- For plain lists of words. The 2019 meta-analysis found a negative effect there.
It also has a cost in feeling. Scores during mixed practice are lower, and people rate it as less effective. Expect that, and judge by the later test.
How Axo uses it
A Bite teaches one idea at a time, because new material needs that. The mix comes after. A unit check draws ten mixed questions from the unit. A capstone mixes every unit. Keep Sharp mixes ideas from different units and different Paths in one session.