Key evidence & numbers
| Number | Where it comes from | What it does not establish |
|---|---|---|
| Under 100 ms | The score being questioned | That the cue caused the response |
| 110 ms | Lower response window in one visual study | A universal limit for every human response |
| 231 ms | Mean measured simple visual reaction time in that study | A target or percentile for your browser |
Woods and colleagues used unpredictable stimulus timing, practice trials and a response window to define which events counted. These choices are part of measuring a response to the signal rather than a premature or mistimed response. The paper also corrected its results for equipment delays, showing why a software timestamp cannot simply be equated with a physiological event. Woods et al., 2015
Reaction and prediction are different behaviours
If the next event is predictable, movement can be prepared before it occurs. A response near the expected time can therefore be extremely close to the event without being initiated after its detection. That is not the same question as simple reaction time to a newly appearing, unpredictable cue. Task instructions and stimulus schedules are essential to interpreting an unusually small interval. Woods et al., 2015 Woods et al., 2015
A choice response adds selecting an action. The classic response-alternative literature shows that task structure affects latency; a minimum claimed for a simple visual game should not be transferred to a game requiring decisions among several actions. Hick, 1952
Why a single spectacular score is weak evidence
The fastest observation is a tail value. Repeating a task many times gives more opportunities for an unusual observation, an anticipatory response or a measurement error. Research on reaction-time outliers explains why investigators define treatment of extreme responses instead of treating every displayed number as equally informative. Ratcliff, 1993
What affects it?
Device precision is not physiological validation
Browsers can record fine-grained timestamps, but timestamp resolution is not the same as accuracy about physical stimulus onset and input. Platform comparisons document both offsets and variability. A score with several decimal places is not necessarily a more accurate estimate of your response to the visible cue. Anwyl-Irvine et al., 2020 Bridges et al., 2020
Nor is it valid to use an equipment correction from a different setup to rescue or disprove a particular result. Calibration applies to the measured apparatus. Without it, the direction and size of all timing errors in your own setup are uncertain. Woods et al., 2015 Anwyl-Irvine et al., 2020
What to do with an unusually fast result
Keep the full run and its rules in view: were there premature responses, were intervals predictable, and was the number a minimum or an average? Compare the same statistic across a fixed number of attempts. Do not increase the number of tries and interpret the new minimum as though the protocol were unchanged. Trial-count research shows that measurement properties depend on the data collected and the question asked. Miller, 2023 Ratcliff, 1993
An isolated score also tells little about a stable individual characteristic. Reliability must be demonstrated, particularly when a task is used to rank people. The appropriate conclusion is that an unusual event occurred in this game; confirming an exceptional human response requires a controlled measurement. Hedge et al., 2017
There is no basis here for calling someone dishonest, diagnosing a problem or defining a universal hard floor. The important distinction is between a displayed interval and evidence that a new sensory event actually triggered the response. Woods et al., 2015 Bridges et al., 2020
Try the related free test
Reaction time test
Try the reaction time test and follow its instructions about waiting for the cue. Interpret the completed run, including premature responses, rather than treating one minimum as a certificate of reaction speed.
Common questions
Is every sub-100 ms score impossible?
These sources do not establish an absolute limit for all response types. They show why an isolated browser score is insufficient evidence of an unpredictable visual reaction. Woods et al., 2015
Why do studies reject very fast responses?
They define valid response windows to reduce contamination from events that do not represent the intended response process. The particular window depends on the experiment. Woods et al., 2015 Ratcliff, 1993
Can I compare a visual score with a sports starting rule?
Only with attention to the different stimulus, movement and measurement system. A rule in one setting is not a calibrated norm for another. Woods et al., 2015 Anwyl-Irvine et al., 2020
Sources
- Woods et al. (2015). Factors influencing the latency of simple reaction time.
- Woods et al. (2015). Age-related slowing of response selection and production in a visual choice reaction time task.
- Hick (1952). On the Rate of Gain of Information.
- Ratcliff (1993). Methods for dealing with reaction time outliers..
- Anwyl-Irvine et al. (2020). Realistic precision and accuracy of online experiment platforms, web browsers, and devices.
- Bridges et al. (2020). The timing mega-study: comparing a range of experiment generators, both lab-based and online.
- Miller (2023). How Many Participants? How Many Trials? Maximizing the Power of Reaction Time Studies.
- Hedge et al. (2017). The reliability paradox: Why robust cognitive tasks do not produce reliable individual differences.
Explore your daily routines in Helibrain
The app has short tests of reaction time, working memory, self-control, reasoning and creativity. Retake them whenever you like, see your trend, and find habits that may help.
Start free in your browser
