Skip to content
One Quick Try

Attentional Blink Test

Letters flash ten a second. Catch the colored one, and see whether an X right after it slips past you.

Takes
about 2 min
Result in
points

Attentional Blink

~2 min

  1. Letters flash in the middle, about ten a second. One is in color.
  2. After each stream, pick the colored letter, then say whether you saw an X after it.
  3. 18 short streams, about two minutes. Keep your eyes on the cross.

Keyboard: type the letter, then Y or N.

Your result: — points

How it's measured
  • Flanker Test

    Which way does the middle arrow point?~1 min
  • Visual Search

    Find the one T hidden among the Ls.~1 min
  • Stroop Test

    Name the ink color, not the word.~1 min
  • Go/No-Go

    React to the circle. Hold back on the square.~1 min
  • Number Flash

    Numbers flash, then hide. Tap them in order.~1-3 min
  • Object Tracking

    Keep your eyes on the marked dots.~2 min

What is the attentional blink?

The attentional blink is a short gap in awareness right after you notice something. When two targets come close together in a fast stream, catching the first makes you miss the second for a moment, as if attention blinked. Jane Raymond, Kimron Shapiro and Karen Arnell named it in 1992: in their study, seeing an X after a target letter became much harder from about 180 to 450 milliseconds after the target.

How does this test work?

Letters flash one at a time in the middle of the screen, each for about 100 milliseconds. One is in color: after the stream, you pick it from four letters. Then you say whether you saw an X after it. In most streams an X comes 1, 2, 3, 5 or 7 letters after the colored one, its lag; in 3 of the 18 streams there is no X at all.

Each letter stays up for a whole number of screen refreshes, counted frame by frame, so every letter lasts exactly as long as the others: six frames on a 60 Hz screen. The page measures your screen first, and your result shows the real timing in milliseconds.

How is your blink worked out?

For each lag, the page counts how often you saw the X, using only the streams where you named the colored letter correctly: if the first letter didn't catch your attention, there was nothing to blink. Your blink is how much more often you saw the X at lag 7 than at lags 2, 3 and 5, in percentage points. A blink of 30 means that, in the moment after catching the first letter, you missed the X 30 points more often than when it came later.

Beside it are your detection at every lag, how often you named the colored letter, and how often you said “yes” when there was no X. A high false-alarm rate makes the other numbers less telling.

Why is an X right after the colored letter often seen?

Many people see the X best at lag 1, straight after the colored letter, and worst a little later. This is called lag-1 sparing: the letter right behind the target seems to get in with it before the gap opens. Troy Visser, Walter Bischof and Vincent Di Lollo reviewed it in 1999 and found it depends on the two targets being in the same place, as they are here. So a blink often looks like a dip, not a slope: fine at lag 1, low at lags 2 to 3, recovering by lag 7.

For other limits of attention, try the Flanker Test or Visual Search.

Questions

What is a normal blink?
It varies a lot between people and with how a test is built, and there is no published average for this one, so the page doesn't give one. Some people show almost no blink.
Why are only some streams counted?
Detection of the X only counts when you named the colored letter correctly, as in the original studies: the blink is what catching the first letter costs.
Does my screen's refresh rate matter?
The page counts whole frames, so each letter lasts the same time on any screen: six frames at 60 Hz, twelve at 120 Hz. The exact time per letter is shown with your result.
Why did I say there was an X when there wasn't?
In a fast stream, letters blur together, and expecting an X makes you more likely to report one. The false-alarm rate is shown so you can see how often it happened.
Is this an ADHD test?
No. It is a demonstration of a normal limit of attention that everyone has, not a diagnostic test.