Interleaving, explained for teachers

Most homework is blocked: ten problems on one skill, then ten on the next. Interleaving mixes them. A set might contain two problems of the first type, one of the second, one of a third, and then another of the first, in no predictable order. It feels more chaotic, and it tends to produce worse performance during practice and better performance later.

Why it works, as far as anyone can tell

When every problem on the page is the same type, the student never has to decide which method applies. They just keep running the one they used on the last problem. On a test, the type is not announced, and choosing the method is half the difficulty. Interleaving makes students practice that choice, and it also spaces the practice of each type, because you cannot return to the first type until you have worked through others. It sits in the family of ideas that psychologist Robert Bjork calls desirable difficulties: conditions that slow learning down in the short run and improve retention.

Where it comes from

Early support came from motor learning and from studies of category learning. One well known experiment by Kornell and Bjork had adults learn to recognize paintings by different artists, either all the paintings by one artist together or mixed. The mixed group was better at identifying new paintings by the same artists, and most participants believed the blocked set had taught them more. In mathematics, Doug Rohrer and Kelli Taylor ran studies in which students practiced finding the volumes of different solids, blocked or mixed, and the mixed group did better on a later test. Rohrer has since run classroom studies in math as well.

A classroom example

A seventh grade teacher has just finished three types of proportion problems: unit rates, scale drawings, and percent change. The blocked homework has eight of each in turn. The interleaved version has the same twenty-four problems shuffled. Students will complain, since any single problem is harder when you cannot tell which kind it is. That friction is the point, so it helps to say so up front.

Common mistakes

• Interleaving before students can do each type at all. Mixing helps consolidate and discriminate. It does not teach a method for the first time. Teach and block first, then mix.
• Mixing things that have nothing in common. The gain comes from types that are easy to confuse. Shuffling unrelated topics together is closer to a cumulative review, which is fine, but a different thing.
• Not telling students. Without an explanation, a harder-feeling set reads as a sign they are failing.
• Judging it by practice scores. Practice accuracy will drop. The test is performance after a delay.

The honest limits of the evidence

The evidence is strongest for mathematics and for tasks that involve telling similar categories apart. It is weaker and mixed for other content, such as reading comprehension or learning a body of facts, and some studies found no benefit or a cost when the material was complex enough that students needed the blocked time to understand it at all. Desirable difficulty cuts both ways: a difficulty is only desirable if the student can actually succeed at it with effort. Heavily interleaved practice for a student who has not yet grasped the basics is just frustration. And much of the classroom research is short, so what happens over a full year of interleaved homework is less well documented.

How this shows up in ClassSense

Honestly, it mostly does not. Practice inside ClassSense is organized by skill pack, so a student works on one skill at a time, which is blocked practice. Mastered skills do come back later for a quick check, which gives some spacing, but that is not the same as shuffling problem types in one sitting. If you want a mixed set, the Create Content tab can generate a worksheet or assessment from your material as a .docx, and you can edit it to blend question types before you print it. The mixing is your design decision.

Common questions