The browser can record when it requested a visual change, but the pixels do not necessarily change at that exact instant. Refresh scheduling and panel processing create delay between software and visible light.
Calibrated reaction-time research often measures and corrects hardware delay when absolute timing accuracy matters.
Mouse switches, keyboard scanning, touch controllers, USB polling and system event handling all add time between the physical action and the browser receiving it.
Different devices can therefore produce different recorded results for the same person.
Suppose someone claims “all monitors add 10 ms” and subtracts 10 from every result. Real devices do not share one fixed delay. Smartphone research alone has measured large differences in combined display and touch latency across models.
A meaningful correction requires measuring the actual device under the actual configuration.
Treat the absolute number as a system-plus-human benchmark and focus on repeatability. If you use the same device, browser, input and display settings, many hardware delays remain relatively stable from session to session.
That makes changes in your own performance easier to interpret even when you cannot perfectly separate biological and hardware components.
Yes. The registered browser time includes delays in displaying the cue and receiving your input.
Only if you have a trustworthy measurement of your specific display path. Generic subtraction can create a false sense of precision.
No. Refresh rate affects frame opportunity, but panel processing, input hardware and software still add latency.