Working memory

Corsi Block Test

Blocks light up in a sequence; you tap them back in the same order. Get it right and the sequence grows by one. Where you stop is your visual span — free, no account, about a minute in your browser.

Take the test

Watch the blocks light up, then tap them in the same order.

READY

0

blocks in order

That is your visual span — how long a spatial sequence you can hold and play back. Much of the movement between runs is attention rather than capacity, so a single number wanders more than you would expect.

No sequence came back complete this time. The first one is only three blocks — worth another go.

Longest correct
Rounds
Grid
3 × 3

Needs sight — the flashes are the only signal. Nothing is uploaded; your span stays in this tab.

What the Corsi block test measures

Your visual-spatial span: how long a sequence of positions you can hold intact long enough to reproduce it. Not how well you remember shapes, and not how fast you are — how much spatial order survives the few seconds between watching and tapping.

The design goes out of its way to keep language out. Nothing on the board has a name, so there is nothing to rehearse under your breath, and the sequence has to be held as a path rather than as a list. That is what separates this from digit span, and why the two can come out differently for the same person.

The ladder does the measuring. Rather than scoring one fixed length, the task keeps growing the sequence until you cannot follow it — so the result is the edge of your capacity rather than a percentage of some arbitrary target.

Where it comes from

  1. 1972

    Corsi builds the spatial twin of digit span

    Philip Corsi, working on his doctorate at McGill, needs a visual-spatial version of the familiar "repeat these digits" test. His answer is a board of nine blocks tapped in sequence — no words, no numbers, nothing that can be rehearsed under your breath.

  2. Then

    Two memories, not one

    The task earned its place because people can be good at one span and poor at the other. That dissociation is a large part of the evidence that verbal and visual-spatial working memory are separate systems rather than one general capacity.

  3. Today

    A clinical fixture

    Versions of the block-tapping task sit in neuropsychological batteries worldwide, and the original nine-block layout is still the standard. What varies is everything else: board arrangement, timing, how many failures end the run.

How to read your result

One number, and it is deliberately blunt: the longest sequence that came back complete.

  • Your span

    The longest sequence you played back correctly, start to finish. Partial credit does not exist here: five of six taps right is a failed round, because a sequence you nearly have is a sequence you do not have.

  • Rounds

    How many sequences you were shown. A high span reached in few rounds means you were not slipping on the way up; the same span after several failures is a shakier number.

  • Grid

    The board it was measured on — 3×3, nine blocks, which is the layout the original test used and the ceiling on how long a sequence can get.

What this version cannot tell you

  • A span is a coarse number

    It moves in whole blocks, so there is nothing between 5 and 6. Two runs differing by one are not meaningfully different, and published figures for adults sit around five or six.

  • Naming the blocks is cheating the task

    Number the positions in your head and you turn a spatial test into a verbal one — a different memory system entirely. It will raise your span and stop measuring what the test is for.

  • Screen and timing matter

    The original used physical blocks tapped by an examiner at their own pace. A fixed on-screen cadence on a phone is not the same instrument, so values here do not map onto published norms.

  • It is not a diagnosis

    Block tapping is used clinically, always alongside other measures and interpreted by someone qualified. A short span on one browser run means you had a short span on one browser run.

Questions about the Corsi block test

What is a normal Corsi span?

Published figures for healthy adults usually land around five or six blocks, which is close to the familiar digit-span range. Treat that as background rather than a target: span is a coarse measure that moves in whole steps, our timing is not the clinical timing, and a run on a phone is not a run in a clinic.

Why does a single wrong tap end the round?

Because a sequence you nearly reproduced is not a sequence you held. Partial credit would let a lucky first half inflate a span, and span is defined as the longest sequence you can play back in full. Two failures at the same length end the run — the first one may be a slip, the second is where you actually are.

Can I count the blocks in my head instead?

You can, and it works, and it stops the test from measuring what it is for. Naming positions moves the job into verbal memory, which is a separate system — the whole reason Corsi built a wordless task in the first place. If you catch yourself doing it, your span here is not a spatial one.

Why does this go up to nine when the app does not?

They are measuring different things. This page is a span test, so it keeps growing the sequence until you cannot follow it — up to the nine blocks the board holds. In the NEURA app, Corsi is a training drill on an adaptive difficulty curve, and the number you see there is a game score feeding your Focus Index, not a span. Nothing on either side contradicts the other; they simply are not the same quantity.

Do I need an account?

No. Everything runs in your browser, nothing is submitted, and your span is shown to you and then gone when you close the tab.

One run is a snapshot. The trend is the point.

A span from one afternoon moves with how alert you happened to be. In NEURA the drill adapts to you and its result joins thirteen other games in a single Focus Index, so what you watch is your own line over weeks — the part no single page can show.