Reaction time is the delay between a signal and your response to it. It is one of the oldest measurements in experimental psychology. The Dutch physiologist Franciscus Donders used it in 1868 to measure how long mental steps take, and it is still used today in sports science, ergonomics, neurology and the design of interfaces.
What typical numbers look like
For a simple task, such as pressing a key as soon as a screen turns green, healthy young adults average around 200 to 250 milliseconds for a visual signal. Sound is faster: about 150 to 190 milliseconds is typical, because the auditory route to the brain is shorter. Touch is in between. These figures include everything between the stimulus and the key press, and part of it is the equipment: a slow screen, a wireless mouse or a touch panel can easily add 30 to 100 milliseconds. A reaction time measured in the browser is therefore a combination of you and your device, and comparing yourself with other people on other devices is not very fair. Comparing yourself with yourself on the same device is.
The time is spent on several stages: the signal is detected by the eye and travels to the brain (tens of milliseconds), the brain recognises it and decides to act (the largest part), and the signal travels to the muscles and the muscles contract. In athletics, a start is called false if a runner reacts within 100 milliseconds of the gun, as it is assumed that this is faster than a reaction can happen and that the runner must have anticipated the sound.
Choice makes it slower: Hick's law
If you have to choose between several responses, the time goes up. In 1952 William Hick and Ray Hyman showed that choice reaction time grows with the logarithm of the number of alternatives: roughly it is a constant plus a factor times log2(n + 1). Going from one option to two adds noticeably more than going from seven to eight. This is why menus with a few well-grouped options are quicker to use than one long flat list, and why a pilot's most urgent controls are limited and distinct.
Speed and accuracy trade off
Nobody can be as fast and as careful as possible at the same time. If you want fewer mistakes, you have to respond more slowly; if you hurry, errors go up. This is the speed-accuracy trade-off, and it is the reason why a tap game with only a reaction time would be a poor test. The Go/No-Go game puts it to work: you must tap on green and hold back on red. A player who taps as fast as possible always will, sooner or later, tap on red too. The best results come from a steady, slightly cautious rhythm.
Researchers describe this with drift-diffusion models, in which evidence accumulates until it reaches a threshold. A high threshold gives slow and accurate responses, and a low threshold gives quick, error-prone responses. People set the threshold according to the situation and the instructions, usually without noticing.
What changes your reaction time
- Warning signals. A cue that warns you something is coming speeds you up. The unpredictable signal is slower.
- Fatigue and lack of sleep. Both increase the average and, even more, the number of very slow responses ("lapses"). Reaction time tasks are used to detect drowsiness for this reason.
- Age. Simple reaction time is shortest in the twenties and increases slowly with age, by a few milliseconds per decade. Practice and experience partly make up for it, since experienced people anticipate better.
- Practice. A few sessions bring a clear improvement on a specific task, especially in the first minutes. After that, improvement is small and mostly about consistency.
- Distractions and stimulants. Caffeine can speed responses slightly. A phone call or text message slows them a lot.
Anticipation versus reaction
There is a pitfall: if you know when the signal will come, you can press at the moment you expect it, and that is no longer a reaction. In a good test the waiting time is random, so that you cannot guess. Our Reaction game does this, and a tap before the green signal scores zero. The Timing and Rhythm games test the opposite skill: not responding as fast as possible but hitting a moment you can predict. They are a nice contrast, since both use the same hand and very different parts of your brain.
How to get your best result
- Sit still, with your finger ready, and look at the middle of the area where the signal appears.
- Do not try to guess. Wait for the signal.
- Use the same device for comparisons, and discard the first try.
- Look at the average of several rounds, not the best one. The median is more stable than the best time, which is partly luck.