Chunk information before asking working memory to use it

Master the parts before combining them so assembly fits within working memory.

Why it works

Working memory can hold roughly four to seven chunks, but a "chunk" can be arbitrarily large if it is already consolidated in long-term memory. Experts can perform complex operations that would overwhelm a novice because each move is a single practiced chunk, not a string of elements. Teaching or practicing component skills to automaticity before combining them grows effective working-memory capacity without exceeding its actual limit.

How to do it

1. Identify the sub-skills a complex task requires and master each until it is automatic.
2. Test that a component is truly automatic: it should consume no noticeable attention.
3. Only then combine components, starting with the two most interdependent.

Evidence

Chunking is one of the most robust findings in cognitive psychology, originating from Miller’s "magical number 7" paper and elaborated through decades of expertise research. Expert chess players encode familiar board positions as single chunks, freeing working memory for strategy. (observational)

Miller’s "7 ± 2" figure is a rough average; modern estimates cluster around 4 chunks. The key insight is that chunks must be automatized, not just reviewed.

Sources

• Miller (1956), "The magical number seven, plus or minus two", Psychological Review
• Chase & Simon (1973), expert vs. novice chess perception, Cognitive Psychology

Common mistake

Combining skills before either is automatic, so both compete for limited resources and the combined task degrades both components simultaneously.