WebToolArcWhy median of five?
One distracted trial can drag a mean upward. The median keeps a short series from being dominated by one lapse while the full trial list still exposes the spread.
Home / Speed & Input Tests / Reaction Time Test
Speed & Input Tests
Reaction Time Test
Run a five-trial visual reaction benchmark with false-start rejection, paint-aligned timing, median and spread statistics, device-context evidence, moving-target practice, local run comparison, and exportable results.
five-trial browser benchmark · local only
Paint-aligned protocolReaction Time & Device-Latency Studio
Use a paint-aligned visual signal, reject false starts, score the median of repeated trials, inspect spread and refresh-frame evidence, then compare saved runs only under compatible conditions.
No result is a medical, neurological, or hardware-latency certification.
—Median · headline
—Mean
—Best
—Worst
—Sample SD
—IQR
0 / 5Valid trials
0False starts
| Trial | Time | Response | Frame evidence |
|---|
Local run history
Up to 12 runs stay in this browser. Scores are never added to the settings-only share URL.
A/B run comparison
Select two saved runs to compare.
The page labels runs compatible only when mode, trial count and declared input profile match. Even then, browser load and hardware conditions may differ.
Why paint-aligned timing?
The random timeout decides when to request the signal, not when to start the stopwatch. The studio crosses animation-frame boundaries before starting the response timer so a JavaScript state change is not treated as visible light.
Simple and moving-target are different
Simple mode measures detection plus response. Moving-target practice adds visual search and pointer movement, so its scores should not be compared directly with classic simple reaction benchmarks.
Use repeated trials instead of one lucky click
The default protocol scores five valid responses and uses their median as the headline result. False starts are rejected rather than rewarded, while the table keeps every valid trial visible so you can inspect consistency instead of hiding it behind one number.
The browser measures you plus the device chain
A web reaction test includes display refresh, panel processing, browser scheduling and input-device latency. The studio estimates the animation-frame interval for context, but it does not subtract that estimate from your score because the monitor and input path contain additional delays the page cannot observe.
Compare like with like
Local A/B comparison is most useful on the same device, browser, mode, trial count and input profile. Cross-device numbers can differ even when the same person reacts equally fast.
Moving-target mode is practice, not the same benchmark
When the signal appears at a random position, the measurement includes visual search and cursor movement. Keep it separate from the simple full-area test used by classic reaction-time benchmarks.
Not a medical or neurological assessment
This browser utility is for repeatable local timing practice and device-context comparison. It cannot diagnose reflex disorders, attention conditions or neurological function, and it is not laboratory instrumentation.