What the reaction time test asks you to do
The practice here shows a waiting screen. After an unpredictable wait of 2 to 8 seconds, the signal appears, and you press as fast as you can: anywhere on the screen, or the space bar. A press during the wait is a false start and costs that trial.
You choose how many trials to run, from 6 to 13. The result is the median of your valid responses, plus a count of false starts and a count of lapses.
What it measures, and what it does not
It measures simple reaction time. That is how long it takes to see one signal you were already waiting for and make one movement you had already decided on. Eyesight, attention and the speed of your hand all go into it.
It does not measure how well you think, decide or remember. A single round is also a small sample. A notification, a sneeze or a slow screen can move your median further than any real change in you would.
Your device is part of the number. A finger on glass, a mouse button and a key each add a different delay, so a number from one input cannot be read against a number from another. This site compares your round only with rounds played on the same kind of input, and it names the pool it used.
Where the numbers come from
The 500 ms line this practice draws is the standard one. The practice borrows its scoring rules from the psychomotor vigilance test, a 10 minute research task in which a lapse is a response of 500 ms or more (Basner, Mollicone and Dinges, Acta Astronautica, 2011). A round here is much shorter, 6 to 13 trials. The same paper suggests a lower line, 355 ms, for its 3 minute version. This practice keeps 500 ms so that rounds stay comparable.
The same paper treats a response under 100 ms as a false start rather than a reaction, because it arrived too early to be a reply to the signal. This practice uses both rules, so a very early press is thrown out rather than counted as your best trial.
That paper also reports that sleep loss slows response times and raises the number of lapses. So compare each round with your own earlier rounds on the same device. Try one rested and one tired.
How to practice it here
Run a short set first to get used to the wait, then a longer one. Keep the same device and the same input every time, or the numbers are not comparable. Watch the lapse count as well as the median, because lapses move first when attention drops.
Start with the reaction time practice. Its short lesson, Wait for the signal, covers how to hold a steady readiness without jumping early. The rest of the timed tasks are on the Attention page.
Related practices on this site
- Stroop test. The same kind of timing, with a conflict added.
- Visual search. Find one named shape among distractors, against the clock.
- Symbol coding practice. Copy a symbol key quickly and accurately.
Questions people ask
What does a reaction time test measure?
It measures simple reaction time: the milliseconds between a signal you are waiting for and one movement you have already decided on. Eyesight, attention and the speed of your hand all go into it. It does not measure how well you think, decide or remember. Your device adds its own delay, so compare rounds played on the same device.
What counts as a slow response, or a lapse?
In the standard 10 minute psychomotor vigilance test, a lapse is counted as a response of 500 ms or more. Basner, Mollicone and Dinges give that standard in a 2011 paper, and suggest a lower line, 355 ms, for a 3 minute version. The same paper treats a response under 100 ms as a false start, because it arrived too early to be a reply to the signal. The practice on this site uses both rules.
Why is my reaction time different on my phone and my computer?
Each input adds a different delay. A finger on glass, a mouse button and a key are not the same, so a number from one input cannot be read against a number from another. This site compares your round only with rounds played on the same kind of input, and it names the pool it used. Keep the same device when you compare your own rounds.
What makes reaction time slower?
Sleep loss is one known cause. Basner, Mollicone and Dinges report that it slows response times and raises the number of lapses. Small things matter too. A notification, a sneeze or a slow screen can move the median of one round further than any real change in you would. So compare each round with your own earlier rounds on the same device.
How does the reaction time practice here work?
You choose from 6 to 13 trials. Each trial waits an unpredictable 2 to 8 seconds, then shows the signal, and you press anywhere on the screen or the space bar. A press during the wait is a false start and costs that trial. The result is the median of your valid responses, with a count of false starts and a count of lapses.