In one sentence
When learning categories that can easily be confused, alternating between them may improve classification for children and young adults, but learners of different ages may read their own progress differently.
What the researchers did
Dong and colleagues compared interleaving (mixing examples from different categories) with blocking (studying one category at a time) while children and young adults learned to identify natural rock categories. The question was not only whether they could classify the rocks later, but also whether their confidence matched their performance.
In the first experiment, participants studied in one of the two schedules and then made an overall judgment about how well they had learned. In the second, they controlled their own study time and rated their learning item by item during study. This allowed the researchers to look at category performance and self-monitoring in two settings.
What they found
- Across both experiments, interleaving produced higher categorization accuracy than blocking.
- Young adults benefited more from interleaving than children did.
- For children, interleaving reduced overconfidence: their judgments came closer to their later performance.
- For young adults, interleaving was associated with underconfidence in both overall and item-by-item judgments.
- The same study schedule can therefore affect accuracy and confidence in different ways.
What this means for learners and educators
- When examples are similar enough to mix up—such as species, artworks, problem types, or rock samples—try alternating categories rather than finishing all examples of one category first.
- Do not treat a feeling of difficulty as proof that the schedule is failing. Mixing can feel less fluent while still supporting later classification.
- For younger learners, brief checks that ask them to identify an example may be more informative than asking, “Do you feel ready?”
Limitations and what we don't know yet
The abstract does not report the sample sizes, the exact age ranges, or how long learning lasted, so the scale and durability of the effects cannot be judged from this report alone. The materials were natural rock categories, not ordinary school units; results may differ for vocabulary, mathematics, or history. The study also shows different confidence patterns by age, but it does not establish why those patterns occur or which classroom supports would calibrate confidence best.