Corsi Block-Tapping Test

Philip Corsi built the first version at McGill in the early 1970s. Nine identical wooden blocks sat at scattered positions on a board. He tapped out a sequence and the volunteer tapped it back.

The scattered layout is deliberate. Blocks in a neat grid can be named by row and column. Once you can name them you can say them to yourself, and the test quietly turns into a word test.

Corsi worked alongside Brenda Milner, who studied the patient known as H.M. Their striking early finding was that block span and digit span could come apart. Damage could wreck one and leave the other untouched. That split is why the task still exists fifty years later.

Words have a trick that positions do not. You can say a list of digits to yourself over and over, refreshing it faster than it fades. Now try that with 'third from the left, then bottom right, then middle'. There is no inner voice for places, and holding them in mind eats the attention the sequence itself needs.

So block spans usually land one to two units below digit spans in the same person. Comparing your result here with your digit span tells you more than either number alone. How working memory works explains the two-system idea that predicts the gap.

Two things are lost on a screen. The real board is solid and you tap it with your hand, so your own movement helps you remember. And a real tester taps at a human rhythm rather than exactly once a second. Both changes make the screen version a little easier for some people and a little harder for others. Neither can be corrected.

What survives is the shape. This version uses the classic scattered nine-block layout rather than a grid, sized to fit the canvas, with blocks that never move between rounds. No sequence taps the same block twice in a row. Your taps only count once the sequence has finished playing, so an early tap cannot spoil a round.

Published figures, most usefully from Kessels and colleagues in 2000, put healthy adult block spans around five to six, with a fairly narrow spread. Do not drop a browser score straight into those tables. The speed, the screen size and the mouse all differ from the standard procedure. For the word-based comparison, what counts as a good digit span has the percentile picture.

Use your score as a lap time instead. Run it on the same device at the same time of day. Treat a change of one block as noise until several runs agree.

One last note on what this is not. It asks you to remember where things were. It does not ask you to rotate a shape or find your way anywhere.