Evidence for Chunking #

Every substantive claim on the Chunking page is checked against current research. Here is each claim, how well today’s evidence supports it, and the sources. The full, de-duplicated source list lives on the references page.

Supported · strong evidence — Working memory is a small, limited-capacity store for information held and manipulated in the present moment, distinct from long-term memory.

The existence of a limited-capacity working-memory store distinct from long-term storage is a foundational, well-replicated finding in cognitive psychology; Cowan (2010) reviews the extensive evidence.

Sources: Cowan, N. (2010), The Magical Mystery Four: How Is Working Memory Capacity Limited, and Why? — Current Directions in Psychological Science, 19(1), 51-57. https://doi.org/10.1177/0963721409359277 · full reference ›

Supported · moderate evidence — When rehearsal and covert grouping are controlled, the capacity of working memory for independent chunks is about four (roughly three to five), not seven.

Cowan (2010, and earlier 2001 review) synthesises evidence that the pure capacity limit is ~4 chunks once participants are prevented from rehearsing or chunking; this ~4 estimate is the current mainstream view, though the exact number and whether capacity is best modelled as a fixed slot count or a graded resource remain debated.

Supported · moderate evidence — Miller’s 1956 ‘magical number seven, plus or minus two’ overstated raw working-memory capacity, partly because participants were already chunking items.

Miller (1956) himself noted the role of chunking; later work (Cowan 2001, 2010) attributes much of the apparent span of ~7 to grouping and rehearsal, leaving a smaller pure capacity of ~4. The reinterpretation is widely accepted.

Sources: Cowan, N. (2010), The Magical Mystery Four — Current Directions in Psychological Science, 19(1), 51-57. https://doi.org/10.1177/0963721409359277 · Miller, G. A. (1956), The Magical Number Seven, Plus or Minus Two — Psychological Review, 63(2), 81-97. · full reference ›

Supported · strong evidence — Working memory capacity is measured in chunks rather than in raw items, so grouping several items into one chunk lets more information be held at once.

That the relevant unit of working-memory capacity is the chunk, and that chunking increases the effective amount of information retained, traces to Miller (1956) and is consistently supported in modern reviews (Cowan 2010).

Sources: Cowan, N. (2010), The Magical Mystery Four — Current Directions in Psychological Science, 19(1), 51-57. https://doi.org/10.1177/0963721409359277 · full reference ›

Supported · strong evidence — Chunks are formed by associating items with knowledge already held in long-term memory, so greater domain expertise enables larger chunks (e.g. expert chess players recall board positions as a few familiar formations).

The chess expertise work originating with Chase and Simon (1973) and extended by Gobet and colleagues shows experts encode positions as larger, knowledge-based chunks; this is a classic, well-replicated demonstration that long-term knowledge enlarges effective working-memory chunks.

Sources: Gobet, F., Lane, P. C. R., Croker, S., Cheng, P. C-H., Jones, G., Oliver, I., & Pine, J. M. (2001), Chunking mechanisms in human learning — Trends in Cognitive Sciences, 5(6), 236-243. https://doi.org/10.1016/S1364-6613(00)01662-4 · full reference ›

Supported · moderate evidence — Grouping a long list into a small number of meaningful named categories makes it easier to recall than the same items as one undivided list.

Organising material into meaningful categories aids recall is a robust finding from the memory-organisation literature (e.g. Bower et al. 1969 on hierarchical clustering) and is consistent with the chunking account in Cowan (2010); the presentation example illustrates rather than tests this.

Sources: Cowan, N. (2010), The Magical Mystery Four — Current Directions in Psychological Science, 19(1), 51-57. https://doi.org/10.1177/0963721409359277 · Bower, G. H., Clark, M. C., Lesgold, A. M., & Winzenz, D. (1969), Hierarchical retrieval schemes in recall of categorized word lists — Journal of Verbal Learning and Verbal Behavior, 8(3), 323-343. · full reference ›

Supported · moderate evidence — Because working memory is limited, presenting information in chunks lowers cognitive load and frees capacity for thinking and learning.

Cognitive load theory holds that instruction should respect the narrow capacity of working memory; chunking and good organisation reduce extraneous load and improve learning. The theory is widely supported, though effect sizes vary by task and learner expertise.

Sources: Sweller, J., van Merrienboer, J. J. G., & Paas, F. (2019), Cognitive Architecture and Instructional Design: 20 Years Later — Educational Psychology Review, 31, 261-292. https://doi.org/10.1007/s10648-019-09465-5 · full reference ›