Chunk related elements before presenting sequences

Group tightly related pieces of information into single named units before teaching the steps that connect them.

Why it works

Working memory processes chunks, not individual elements. When two or more elements always appear together and must be understood together, treating them as one chunk costs only one working memory slot rather than two or more. This is the mechanism behind why experts learn new material in their domain faster: they have pre-formed schemas (chunks) that compress many elements into single high-level units, leaving more capacity for the new connections.

How to do it

1. Before teaching or learning a topic, identify which concepts must always be understood together to be meaningful.
2. Name each cluster as a single unit ("photosynthesis" rather than "light + water + CO2 → glucose + oxygen as separate items").
3. Ensure the learner can fluently retrieve each chunk before introducing sequences that connect chunks.
4. Test chunk fluency with low-stakes retrieval before building complexity.

Evidence

Chunking is grounded in Miller’s (1956) classic "magical number 7" paper and extensively built on in cognitive load research. The schema-automation mechanism is well supported by studies of expert-novice differences in working memory use. (rct)

The optimal chunk size depends on the domain and the learner’s prior knowledge; over-chunking can hide necessary distinctions from novices.

Sources

• Sweller, Ayres & Kalyuga (2011), Cognitive Load Theory, Springer
• Miller (1956), The magical number seven, Psychological Review

Common mistake

Presenting component elements before establishing what they collectively mean, forcing learners to hold unrelated pieces in working memory until they can see the point of each one.