Key evidence & numbers
| Example | What changes | What does not follow |
|---|---|---|
| Grouping 12 digits into 4 groups of 3 | The representation becomes grouped | Everyone can remember that sequence |
| Recognising a familiar pattern | Existing knowledge can reduce encoding demands | Every symbol occupies an independent slot |
| The trained 7-to-79-digit case | Highly practised digit recall | A 79-item general working-memory capacity |
The first row is an illustrative grouping, not a measured capacity. The third is the original intensive-practice case. In that study, the participant used knowledge of running times and an organised retrieval structure. The paper also reports lack of transfer when the material changed to consonants, which is central to interpreting the striking digit result. Ericsson et al., 1980
What is a chunk?
A chunk is an encoded unit, not necessarily a fixed number of visible characters. Its usefulness depends on what the learner already knows and how the sequence is structured. Miller's discussion of recoding and Mathy and Feldman's work on compression both show why raw item counts can be misleading when structure differs. Miller, 1956 Mathy et al., 2012
Cowan's smaller-capacity account examines conditions that constrain rehearsal and long-term knowledge. A person using meaningful groups in an ordinary digit game is not necessarily performing under those restrictions. The two results need not contradict each other because their units and allowed processes differ. Cowan, 2001
Is grouping the same as repeating?
Rehearsal keeps material active; grouping changes how material is represented or organised. They can work together, but they are not identical operations. Working-memory theory separates temporary maintenance from other control processes and from the contribution of long-term knowledge. Baddeley, 2012
What affects it?
How much improvement should you expect?
The 7-to-79 result is one unusually intensive case, not an average treatment effect or a training schedule for everyone. It provides proof that learned organisation can greatly change a task score, while giving little basis for forecasting an individual's eventual span. The lack of broad transfer is as important as the record. Ericsson et al., 1980
Published span scores also depend on the stopping rule. In the large auditory study, maximum achieved span and traditional span gave different means in the same participants. A strategy comparison must therefore keep presentation and scoring constant, rather than mixing the longest isolated success with a differently measured baseline. Woods et al., 2010
What is reasonable to try?
For an informal comparison, decide whether you will group digits, use the same strategy across comparable runs and record the change if you adopt a new method. The compression research supports the role of structure; it does not validate a universal grouping size or guarantee that a particular mnemonic will help every sequence. Mathy et al., 2012
Use unfamiliar material when judging whether a strategy generalises beyond a memorised example. Span-methods research emphasises matching the procedure to the construct being measured. A learned skill is useful, but it should be described as that skill rather than relabelled as a global capacity increase. Conway et al., 2005
Finally, there is no validated conversion from a chunked digit score to an n-back level or IQ. The meta-analysis comparing span and n-back demonstrates that even tasks sharing the working-memory label are not interchangeable. Redick et al., 2013
Try the related free test
Number memory test
Try the number memory test and note your grouping strategy. If you compare methods, keep the display and scoring rules consistent. A higher score shows better recall under those conditions; it does not certify a general increase in intelligence.
Common questions
Does chunking increase the number of memory slots?
It can change the information represented by a unit. The evidence does not require a matching expansion of a fixed number of general slots. Cowan, 2001 Mathy et al., 2012
Will I reach 79 digits?
The original result was one intensively trained participant, not a typical outcome or promise. Ericsson et al., 1980
Why do familiar numbers seem easier?
Existing knowledge can help recode and retrieve a sequence. That is part of what chunking research examines. Miller, 1956 Mathy et al., 2012
Sources
- Woods et al. (2010). Improving digit span assessment of short-term verbal memory.
- Cowan (2001). The magical number 4 in short-term memory: A reconsideration of mental storage capacity.
- Miller (1956). The magical number seven, plus or minus two: Some limits on our capacity for processing information..
- Baddeley (2012). Working Memory: Theories, Models, and Controversies.
- Conway et al. (2005). Working memory span tasks: A methodological review and user’s guide.
- Redick et al. (2013). Complex span and n-back measures of working memory: A meta-analysis.
- Ericsson et al. (1980). Acquisition of a Memory Skill.
- Mathy et al. (2012). What’s magic about magic numbers? Chunking and data compression in short-term memory.
Explore your daily routines in Helibrain
The app has short tests of reaction time, working memory, self-control, reasoning and creativity. Retake them whenever you like, see your trend, and find habits that may help.
Start free in your browser
