Science & Measurement Notes

Online reaction-time testing can be useful for repeated personal benchmarking, but absolute timing depends on the full device stack.

Browser timing is real, but not hardware-free

Modern browsers provide high-resolution monotonic timing APIs and frame callbacks that make consistent in-browser measurement possible. At the same time, the moment a display physically changes and the moment an input device physically actuates are not directly observable from ordinary JavaScript.

Why external hardware matters in validation studies

Studies that estimate total system timing error commonly use external instrumentation such as photodiodes or other light sensors for actual stimulus onset and controlled physical input devices for response timing. That is why ReactionMetric does not manufacture an absolute “corrected” reaction time.

Best use

Use ReactionMetric to compare repeated sessions under similar conditions, track personal trends and, once the dataset is large enough, compare against transparent browser-observed cohorts.

Primary references

ReactionMetric's timing implementation is based on browser primitives documented by MDN: performance.now() for a high-resolution monotonic clock and requestAnimationFrame() for work aligned with display refresh.

For touchscreen and keyboard browser timing, see Pronk et al., Mental chronometry in the pocket? Timing accuracy of web applications on touchscreen and keyboard devices. The study used external chronometry to quantify stimulus and response timing across consumer devices. For a broader comparison of online experiment platforms, browsers and devices using an opto-detector and robotic actuator, see Anwyl-Irvine et al., Realistic precision and accuracy of online experiment platforms, web browsers, and devices.

Community benchmark pages will always separate literature-derived context from ReactionMetric's own first-party browser data, and will show sample size, collection window, task version, and cohort definition.