Back to Blog

Chunking Memory Technique: Examples and How to Use It

The chunking memory technique groups facts so working memory can hold more. See real examples, why Miller's 7±2 is outdated, and how to chunk your notes.

Rohit SalujaFounder, Active Recalling12 min read
A wall covered in hundreds of colorful sticky notes, a picture of information before it is chunked

The chunking memory technique means grouping separate pieces of information into a few larger, meaningful units, so your working memory juggles three or four chunks instead of a dozen loose items. Try the digits 1 9 4 5 1 9 8 9 2 0 0 1. As twelve digits, they are hard to hold. As the years 1945, 1989, and 2001, they are three items you already know.

That last part is the key, and most explanations skip it. Chunking isn't just cutting information into smaller pieces. It works because each chunk points to something already stored in your long-term memory. The more you know about a subject, the bigger your chunks can be.

This guide covers what chunking is in psychology, why Miller's famous "7 plus or minus 2" is out of date, what experts teach us about chunks, and a step-by-step way to chunk your own course material.

What is chunking in psychology?

In psychology, a chunk is a familiar unit that your memory treats as a single item, and chunking is the process of recoding several items into one. The idea comes from George Miller's 1956 paper "The Magical Number Seven, Plus or Minus Two," which argued that short-term memory is limited by the number of chunks, not by the amount of information in them.

A few everyday examples:

  • Letters into words. C, A, T is three letters for a child learning to read and one word for you.
  • Digits into numbers you know. A phone number is easier as area code, prefix, and line number.
  • Moves into patterns. A chess master sees a familiar castled king position, not six separate pieces.
  • Steps into phases. The ten steps of glycolysis become an "energy investment" phase (steps 1 to 5) and a "payoff" phase (steps 6 to 10).

What counts as one chunk depends on who is looking. "NaCl" is one chunk to a chemistry student and four characters to a toddler. That's why chunking is a skill you build in each subject, not a trick you apply once.

Is Miller's law of 7 plus or minus 2 still true?

Not as a capacity limit. In 2001, Nelson Cowan reviewed decades of studies and concluded that when people can't rehearse or regroup items, working memory holds about four chunks (a range of three to five). He argued that Miller's seven was meant more as a rough estimate and a rhetorical device than as a real limit.

So why do people reliably remember about seven digits? A 2012 study by Fabien Mathy and Jacob Feldman offers an answer. When they built digit lists with hidden patterns, people's spans depended on how many distinct chunks were left after compression. The true limit came out at about 3 or 4 chunks, which is roughly equal to 7 ordinary items, because typical lists contain some patterns people group without noticing.

The practical takeaway: plan for about four things at once, not seven. If a lecture slide asks you to hold eight new terms in mind while following an argument, you will lose some of them unless you group them.

How does chunking help working memory?

A chunk works by pulling a compact pattern out of long-term memory to stand in for all its parts. In a 2019 series of experiments, Mirko Thalmann, Alessandra Souza, and Klaus Oberauer found that remembering a familiar chunk also improved recall of other, unrelated items held at the same time. The chunk took up less room, which left space for the rest.

Two details from the same experiments matter for studying. Bigger chunks came free only when their parts were unique; chunks built from overlapping parts took up more room. And chunks helped most early in a list, where they freed space for the items that came after them.

What to do differently:

  • Make each chunk come back as one unit. A grouping only saves space once you can recall it from its name without rebuilding it, so practice it until it does.
  • Keep groups distinct. Groups that share members blur together. Sort by a property that puts each item in exactly one place.
  • In a brand-new subject, write things down. You don't have many chunks yet, which is why working memory feels so much tighter there than in a subject you've studied for years.

What do experts teach us about chunking?

The classic evidence comes from chess. In 1973, William Chase and Herbert Simon showed three players (a master, a strong club player, and a beginner) a chessboard for five seconds, then asked them to rebuild it from memory. On real middle-game positions with 24 to 26 pieces, the master placed far more pieces correctly on the first try.

Pieces placed correctly after a 5-second look

First-trial recall of real middle-game chess positions (24 to 26 pieces on the board)

Approximate first-trial figures from the paper, rounded. One player per skill level, so read this as a case study, not an average across many players. On randomly arranged boards, the master's advantage disappeared.

Source: Chase & Simon (1973), Cognitive Psychology

View data
PlayerPieces placed correctly
Master16 pieces
Class A player8 pieces
Beginner4 pieces

When the same pieces were scattered at random, the master's advantage vanished in that study. He wasn't seeing more pieces. He was seeing familiar groups of pieces, each one a chunk learned from years of play. Later work by Fernand Gobet and Herbert Simon (1996) found that strong players keep a small edge even on random boards, but the gap is much smaller than on real positions.

Memory span itself can be trained the same way. In a 1980 study, a student known as SF practiced for more than 230 hours and raised his digit span from 7 to 79 digits. He was a keen runner and recoded digit groups as running times, building larger and larger chunks. The gain came from his coding system, not from a bigger memory.

Two lessons for students:

  1. Expert memory is mostly knowledge. The fastest way to "remember more" in a subject is to learn its common patterns, which are stored in long-term memory in explicit and implicit forms.
  2. Chunks are specific. SF's system worked for digits. A chess master's chunks don't help with anatomy. You build chunks one subject at a time.

Chunking examples for students

SubjectRaw itemsChunked version
History12 dates in a unit3 eras, each with a cause and an outcome
Biochemistry20 amino acids4 groups by side chain: nonpolar, polar uncharged, acidic, basic
Biology10 steps of glycolysisInvestment phase (1 to 5) and payoff phase (6 to 10)
ChemistryAtoms in sulfate (1 S, 4 O, 2− charge)One unit: SO₄²⁻
Matha² + 2ab + b²One pattern: "perfect square," recognized on sight
LanguagesJe + voudrais + un + caféOne phrase: "je voudrais…" for ordering anything
CodingIndex, condition, increment, bodyOne idiom: "loop over the list"

Notice what the right-hand column has in common: every chunk has a name or a meaning. A random split ("amino acids 1 to 5, 6 to 10…") would give you smaller lists but no real chunks.

Common mistakes with the chunking memory technique

Splitting without meaning. Breaking a list of 30 terms into three lists of 10 makes it shorter, not chunked. Group by something that's true about the items.

Chunks that are too big. A group of nine is two groups. Add a level instead of stretching one.

Chunking material you don't understand. If you can't explain why items belong together, the grouping won't hold. Understand first, then chunk. Schema theory explains why prior knowledge makes new material easier to organize.

Confusing chunking with mnemonics. An acronym like HOMES (the Great Lakes) is a mnemonic that packs five names into one cue. It's useful, but it's a different tool. See mnemonic devices for when each fits.

Never testing the chunk. Reading neat grouped notes feels productive, but only recalling them from a blank page makes them stick. The method below builds that in.

How to chunk information: a 5-step method

Here's the method applied to a common pain point, the 20 amino acids. Groupings vary slightly between textbooks (tyrosine, glycine, and histidine move around), so use your course's version.

Step 1: Find the structure. Look for a property that sorts the items: side-chain chemistry, in this case. In other subjects, try category, sequence, cause and effect, or function.

Step 2: Make groups of about 3 to 5. Nonpolar (about nine, depending on your textbook), polar uncharged (five or six), acidic (two), and basic (three). If a group is bigger than five, add a level underneath it: split nonpolar into aliphatic and aromatic, for example.

Step 3: Name each group. "Acidic: aspartate, glutamate" is a chunk. The name becomes the retrieval cue you'll use on the exam.

Step 4: Practice recalling the group from its name. Close the book. Write "basic" and list the members (lysine, arginine, histidine). Then go the other way: given "glutamate," name its group and why. Check yourself and fix the misses.

Step 5: Link the chunks. Once each group comes back easily, practice the level above it. Can you sketch all four groups and say which ones sit on a protein's surface in water? Now the whole topic is one chunk with four parts.

Steps 4 and 5 are where most people stop too soon. A chunk you can't retrieve is just a nicely organized page. Retrieval practice turns it into memory, and spaced repetition keeps it there.

Try it tonight: pick one list in your notes with more than eight items (drug classes, causes of a war, irregular verbs). Sort it into named groups of 3 to 5, close the notes, and write out each group from its name alone. Tomorrow, do it again before you look.

Frequently asked questions

What is an example of chunking?

Remembering the digits 194519892001 as three years (1945, 1989, 2001) is chunking. So is learning the 20 amino acids as four groups by side chain, or reading "FBI" as one unit instead of three letters. In each case, several items become one unit because the grouping matches something you already know.

Is Miller's 7 plus or minus 2 still accurate?

Not as a hard limit. Modern research, led by Nelson Cowan's 2001 review, puts working memory at about four chunks when people can't rehearse or regroup. Miller's seven roughly matches how many ordinary items people can hold, because they unknowingly group some of them. Plan for about four new things at once.

How many chunks can working memory hold?

About four, with a typical range of three to five, according to Cowan's review and later studies such as Mathy and Feldman (2012). The size of each chunk varies. An expert's chunk can hold much more information than a beginner's, which is why experts seem to have better memories in their field.

Is chunking a mnemonic?

Chunking is the broader process of grouping items into meaningful units, and it happens naturally as you gain knowledge. Mnemonics such as acronyms are deliberate devices that create a chunk for material with no natural structure. Use real structure when it exists, and a mnemonic when it doesn't.

How do you chunk information when studying?

Find a property that sorts the material (category, sequence, cause, function), make groups of 3 to 5, and name each group. Then close your notes and practice recalling each group from its name, and finally recall how the groups fit together. Review the chunks on spaced days so they stay retrievable.

Sources and further reading

Start with the primary research and official guidance below. Links open in a new tab so you can compare the evidence with the article.

  1. The magical number seven plus or minus two: some limits on our capacity for processing information

    Miller, Psychological Review, 1956

    Read source
  2. The magical number 4 in short-term memory: a reconsideration of mental storage capacity

    Cowan, Behavioral and Brain Sciences, 2001

    Read source
  3. What's magic about magic numbers? Chunking and data compression in short-term memory

    Mathy & Feldman, Cognition, 2012

    Read source
  4. How does chunking help working memory?

    Thalmann, Souza & Oberauer, Journal of Experimental Psychology: Learning, Memory, and Cognition, 2019

    Read source
  5. Perception in chess

    90004-2) — Chase & Simon, Cognitive Psychology, 1973

    Read source
  6. Recall of rapidly presented random chess positions is a function of skill

    Gobet & Simon, Psychonomic Bulletin & Review, 1996

    Read source
  7. Acquisition of a memory skill

    Ericsson, Chase & Faloon, Science, 1980

    Read source