Evidence guide · Measurement
What Does an Online Reaction Time Test Actually Measure?
A 200 ms score is not simply your “brain speed.” It is the measured time across a chain of human processes plus the timing characteristics of the device and browser used to run the test.
The short answer
Your score combines human response time with measurement-system delay.
An online visual reaction test records an interval between a stimulus being presented by the test and your response being registered by the browser. Human reaction time itself involves multiple stages. The recorded browser score can also contain delay from display presentation, input hardware, operating-system scheduling and browser event timing.
This is a conceptual measurement model, not a clinical equation. It is useful because it prevents a browser score from being mistaken for a direct measurement of neural conduction or general intelligence.
What happens between seeing a target and responding?
Reaction time looks like a single number, but the underlying process is not singular. A 2021 clinical neurophysiology review describes reaction-time tasks in terms of stimulus detection and discrimination, decision or response selection, response planning and organization, and response initiation and execution. In a simple reaction-time task, some response preparation can happen before the target appears because the required action is already known.
- Detect the stimulus.Your visual system has to register that the expected target has appeared.
- Interpret what happened.The task determines how much discrimination or decision-making is required. A simple “click when it appears” task requires less selection than a choice task.
- Prepare and initiate the response.The nervous system initiates the already-known movement or selects among possible movements.
- Register the physical input.The mouse, keyboard, trackpad or touchscreen and the surrounding software stack register the action.
This is why “reaction time” and “brain speed” should not be used as synonyms. Reaction time is a behavioral measurement affected by several processing stages and by the way the response is measured. Rothwell et al. (2021) provide a broader neurophysiological review of these stages.
Can a browser measure reaction time usefully?
Yes—with an important qualification. Browser-based reaction-time measurement can detect meaningful behavioral differences, but an absolute browser score should not be treated as laboratory hardware with zero timing error.
≈25 ms
JavaScript offset in one controlled comparison
de Leeuw and Motz compared a JavaScript visual-search task with Psychophysics Toolbox. JavaScript response times were about 25 ms longer, yet the distributions had comparable variability and both systems detected the experimental reaction-time effects.
Read the study3 conditions
Lab software, lab browser and remote web
Hilbig randomly assigned participants to a lexical-decision task in three settings. The expected word-frequency effect appeared strongly in all three, with no evidence that the browser or remote web condition was inferior for detecting that effect.
Read the study200k+
Real-world technology varies
Anwyl-Irvine and colleagues tested online experiment platforms, browsers, operating systems and keyboards, then analyzed technology data from more than 200,000 participants. Their conclusion was nuanced: modern web platforms can provide reasonable timing accuracy and precision, but no platform is best under every condition.
Read the open-access studyWhy the same person can get different scores on different devices
A browser test depends on more than the participant. The screen has to present the visual change, the input device has to register the action, and software has to deliver the corresponding event. Refresh timing, keyboard or pointer hardware, browser choice and operating system can therefore change the measured interval even if the person did not meaningfully change.
That does not make online testing useless. It changes how the number should be interpreted. For personal tracking, comparisons are cleaner when you keep the device, browser, input method and test conditions similar and look at a series of attempts rather than a single unusually fast click.
A 10–20 ms difference measured on two different setups is not automatically evidence that your nervous system became 10–20 ms faster or slower. Hardware and software are part of the measurement system.
Reaction time is not a direct intelligence score
Response time can be useful in cognitive research, but raw response time is a composite outcome. A large 2022 analysis illustrates the problem with treating it as a direct readout of “mental speed.” Researchers modeled data from 1.2 million participants and found that raw response times began slowing from around age 20. Their model attributed the earlier slowing largely to increased decision caution and slower non-decision processes; slowing in the modeled mental-speed component appeared only after roughly age 60.
The important lesson is not that one age threshold applies to every reaction task. It is that a slower raw response can have more than one explanation. The observable number does not identify its own cause. See von Krause, Radev & Voss (2022).
So what does a ReactionTester score mean?
ReactionTester's primary reaction-time game is a browser-based visual-motor task. It is best interpreted as your recorded performance on that task, on that device, under those conditions. It is useful for repeated practice, personal comparisons and community play. It is not a neurological diagnosis, a clinical assessment or a device-independent measurement of pure neural speed.
There is another technical detail worth making explicit. ReactionTester currently timestamps its game state and response in browser JavaScript. As with other web tests, rendering and event-handling behavior can contribute to the interval. We therefore avoid claiming that a displayed score isolates the instant photons reached the retina or the instant muscle activation began.
If you want to compare your own attempts, keep the setup stable and use multiple trials. You can then take the Reaction Time Test and treat the result as a performance benchmark rather than a medical measurement.
Limitations
What this article does—and does not—claim
- This page is a literature review and measurement guide. It does not report a new ReactionTester experiment.
- Studies cited here used different tasks, hardware and populations. Their exact timing offsets should not be copied directly onto every browser test.
- Browser-based tests can detect behavioral effects while still containing systematic hardware or software latency.
- ReactionTester scores are intended for testing and practice, not diagnosis or medical decision-making.
Primary sources
References
- Rothwell J, Antal A, Burke D, et al. “Central nervous system physiology.” Clinical Neurophysiology. 2021;132(12):3043–3083. DOI: 10.1016/j.clinph.2021.09.013. PubMed
- de Leeuw JR, Motz BA. “Psychophysics in a Web browser? Comparing response times collected with JavaScript and Psychophysics Toolbox in a visual search task.” Behavior Research Methods. 2016;48:1–12. DOI: 10.3758/s13428-015-0567-2. PubMed
- Hilbig BE. “Reaction time effects in lab- versus Web-based research: Experimental evidence.” Behavior Research Methods. 2016;48(4):1718–1724. DOI: 10.3758/s13428-015-0678-9. PubMed
- Anwyl-Irvine A, Dalmaijer ES, Hodges N, Evershed JK. “Realistic precision and accuracy of online experiment platforms, web browsers, and devices.” Behavior Research Methods. 2021;53:1407–1425. DOI: 10.3758/s13428-020-01501-5. Open-access full text
- von Krause M, Radev ST, Voss A. “Mental speed is high until age 60 as revealed by analysis of over a million participants.” Nature Human Behaviour. 2022;6:700–708. DOI: 10.1038/s41562-021-01282-7. Nature