What interleaving changes
Most course exercises are blocked: a chapter on weighted averages is followed by ten weighted-average problems. You know the method before you read the question. Dunlosky and colleagues describe interleaved practice as a schedule that mixes different kinds of problems, or different kinds of material, within one session. The point is that the learner has to recognize which approach a problem needs, which is the situation you face at work or in an exam.
Their 2013 review rated interleaved practice moderate utility, noting that its benefits had only begun to be explored systematically. It rated elaborative interrogation (explaining why a stated fact is true) and self-explanation (explaining how new information relates to what you know, or the steps you took) as moderate too, partly because they had not been adequately evaluated in educational settings. Treat these as useful additions, not replacements for practice testing and spacing.
Build a mixed set from existing exercises
Our twelve data-reasoning exercises can be used this way: pick items from different families and attempt them out of order.
- Choose three problem types you have already practised separately, for example weighted averages, percentage change and missing-value handling.
- Take two or three unused problems of each type from the course or your own notes.
- Shuffle them and remove chapter headings or hints that reveal the type.
- Before solving each one, write the method you think applies and one reason.
- After solving, check against the answer key and record both the method choice and the result.
Add a why-question to each item
Write the answer before checking. If you cannot say why a step works, mark the item for a worked-example review, as described in our worked-example guide, rather than moving on.
| After this step | Ask yourself | Example (invented) |
|---|---|---|
| Choosing the method | What feature of the problem told me this? | Groups have different sizes, so a simple average of averages is wrong |
| A calculation step | Why is this step valid here? | I divide by total units, not by number of groups |
| The final answer | Does this make sense given the data? | The result lies between the group values, closer to the larger group |
Worked example: a mixed set of six (invented problems)
A learner has finished separate lessons on weighted averages, percentage change and handling missing values. The mixed set contains two of each, shuffled: a sales total across stores of different sizes, a price that rises and then falls by the same percentage, a survey with blank answers, and three more like these. For each item they first write the method and a reason.
They choose the method correctly on four of six. Both errors are on the percentage-change items, where they averaged the percentages instead of applying them in sequence. That error would not have appeared in a blocked set, because a chapter heading would have told them which method to use. The next session adds two more percentage-change items to a new mixed set.
Common pitfalls
The review summarizes published studies; it does not tell you how much a particular course will benefit. We have not run an interleaving trial. Sources accessed 13 October 2026.
- Mixing types you have never practised alone; start blocked, then interleave.
- Expecting the session to feel easy; mixed practice usually feels harder, which is not evidence it is failing.
- Counting only correct answers; record whether you chose the right method as a separate score.
- Mixing unrelated subjects with no shared decision; interleave problems that could plausibly be confused.
Sources & verification
Product details and prices can change. Check the linked provider before buying.
Sources link directly to providers. Product buttons may use separately labeled affiliate links. Read our disclosure.