Why stopping is harder than going
Reading cannot be switched off and a loaded finger cannot be recalled for free. The attention tests measure the price of overruling yourself.
Most of what a reaction test measures is how quickly you start something. Two of the tests here measure the opposite skill: how well you stop, or override, something you were already about to do. Psychologists call it inhibition, and the name makes it sound like a muscle. It is better pictured as a race between two processes, one that has a head start and one that has to catch up.
A reader that started first
In 1935 John Ridley Stroop published the experiment that still carries his name. Volunteers named the ink color of printed words. When the word was a color name written in a different ink (the word RED in blue), naming the ink took them much longer than naming the colors of plain squares. Later versions, including the test here, compare those words with words that match their ink. The effect has been replicated for ninety years, and the standard review (MacLeod, 1991) counts it among the most robust results in the field.
The usual explanation is a race. Reading a word you know is automatic: it starts without being asked and cannot be declined. Naming an ink color is controlled, and it starts later. When both answers agree there is nothing to resolve, and when they disagree the correct one has to overtake the wrong one before you speak. The delay is interference, and it is the quantity the Stroop test reports: your average time on the words that lie, and the lag, which is the lying average minus the matching average.
That subtraction is more than a convenience. Your screen, your input device and your hand add the same milliseconds to both averages, so they cancel. The lag travels between devices better than almost any other number on this site. It is also why the test uses words in your own language: a word you have to decode slowly interferes weakly, because the reader that is supposed to start first starts late.
A finger that is already loaded
The go/no-go design is older than Stroop’s, and it has a close ancestor in the experiments of Franciscus Donders, whose 1868 method for timing mental stages appears in the science of reaction time: a version in which the subject answers one stimulus and ignores the other. Its modern form is the one in the go/no-go test: most signals ask for a quick tap, a few ask for nothing, and the few are the point.
Making the go signals frequent is the trick. After a run of greens, tapping stops being a decision and becomes a rhythm, so a red has to cancel a movement that is already prepared. Researchers who study this task measure the slip, the tap that escapes on a red (a commission error), and separately the miss, the green that was never tapped (an omission). They tell different stories. A slip is a failure to stop. A miss is a different kind of error, and the two should not be priced as if they were the same thing.
The classic account of how stopping works is the horse-race model of Gordon Logan and William Cowan (1984). A go process and a stop process run independently, and whichever finishes first decides the trial. If the stop signal comes late, or the go process is fast, the go wins and the response escapes. The model has a practical consequence that you can feel in the test: speed and restraint trade against each other. The faster you tap, the less time a red has to catch the movement.
What the two tests are built to do about it
Both machines are built so that trade cannot be exploited. In the go/no-go test, go and no-go share one 1.2 second window, so hand speed buys no extra time to restrain yourself. A slip sends its red back into the lights you have not yet seen and deals extra greens with it, so impulsiveness costs seconds on the clock rather than a deduction computed afterward. In the Stroop test, the score is the time on lying words, the number that cannot be improved by coasting on the easy ones.
None of this is a clinical measure. These are short browser tasks on an unknown device, and a score here says nothing about attention disorders or any other condition. What they can show you is how your own interference or restraint moves from one session to the next, played the same way on the same device. One caution about practice: getting better at the Stroop test is, first of all, getting better at the Stroop test, and a lower lag here is no proof that anything else in your life got easier.
How to read your own numbers
Compare like with like. A rested head and a device you do not switch give a cleaner line than any trick. If your Stroop lag is large but your matching words are fast, you are a quick reader fighting your own fluency, which is the normal case and not a fault.
The curves behind the percentiles, and how many runs the site has measured so far, are declared on the Method page.
Sources
- Stroop, J. R. (1935). "Studies of interference in serial verbal reactions". Journal of Experimental Psychology, 18(6), 643–662.
- MacLeod, C. M. (1991). "Half a century of research on the Stroop effect: an integrative review". Psychological Bulletin, 109(2), 163–203.
- Logan, G. D., & Cowan, W. B. (1984). "On the ability to inhibit thought and action: a theory of an act of control". Psychological Review, 91(3), 295–327.
Keep reading
- What choosing costs: Hick's law and the choice reaction testA simple reaction is a reflex with a clock on it. Add a decision and the time stops being constant, and starts obeying a rule.
- How much a head can holdThe buffer is tiny, the range is narrow, and the only real upgrade ever found is cheating in a way psychologists approve of.