Number Memory Test

Digit span, and the famous number that is usually misquoted

This is a digit span task, one of the oldest measurements in psychology. It is where the popular claim that people can hold seven things in mind comes from — from a 1956 paper by George Miller titled "The Magical Number Seven, Plus or Minus Two".

Miller's own point was subtler than the number it is remembered for. He observed that the limit is not seven digits but roughly seven chunks, and that what counts as a chunk depends entirely on what you already know. That is why the figure moves around so much between people and between tasks.

In practice most adults get somewhere between seven and ten digits on a test like this one, and the tail above that belongs mostly to people who are deliberately grouping rather than to people with unusual memories.

Chunking, rehearsal, and where each one stops working

Two strategies carry nearly everyone past their untrained score. The first is grouping the digits into pairs or threes, so eleven digits becomes four items instead of eleven. The second is rehearsal — saying the number to yourself so it stays in the loop that holds sounds rather than the one that holds shapes.

Rehearsal has a hard ceiling that grouping does not, because the loop holds a couple of seconds of speech. Long numbers stop fitting no matter how quickly you say them, which is the point where people who rely on saying the digits stall while people who group them keep climbing.

Recognising fragments is the strongest trick of all and it is why scores here are not entirely comparable between people. A number containing a year you know, or an area code, or a repeated pair, is genuinely easier for you than for someone who sees only digits.

How long the number is shown, and why that matters

Longer numbers are displayed for longer, in proportion to their length. A fixed display time would turn the later rounds into a reading-speed test, since a nineteen-digit number cannot be read at all in the time a three-digit one needs.

The number disappears before the input appears, so there is no window in which both are visible. That sounds obvious, but it is the detail that decides whether the test measures memory or transcription.

On a phone the keypad is numeric and the input is placed so the on-screen keyboard cannot cover it. An input hidden behind the keyboard is the commonest way a test like this breaks on mobile, and it fails silently — the visitor simply cannot see what they typed.

Reverse is the other button, not a second address

Standard asks you to type the digits as shown. Reverse asks you to type the last digit first. The two share a keypad and a page. They do not share a curve.

People who sail through nine digits forward often drop two or three on reverse. That cost is the reason the scores stay apart. Pooling them would flatter every forward run and punish anyone who only took the reverse button.

The display time still grows with length on both playlists. What changes is only the order you have to emit.

This is a measurement toy, not a medical or psychological test. No score here says anything about your health, attention or cognition, and it is not a screening tool of any kind.

Questions

How many digits should I be able to remember?

Most adults manage between seven and ten on a test of this kind. Beyond that, grouping the digits deliberately is doing most of the work rather than raw memory.

Is it cheating to group the digits?

No, and everyone who scores well is doing it. Chunking is a genuine skill and the reason the same person can improve substantially between their first and fifth attempt.

Why is the number shown for longer on later rounds?

So the test stays a memory test. With a fixed display time the long numbers would be measuring how fast you can read rather than how much you can hold.

Which order do I type the digits?

Standard types them as shown. Reverse types them last-digit-first. Pick the button before you start; switching mid-run resets to idle.