CPSKIT / METHOD

Know what a result means.

CPSKit observes events delivered to a web page. A browser observation is useful evidence within a limited test; it is not a diagnosis of the physical device.

Click speed

The first eligible primary pointer press starts the run and counts as press one. Events are accepted in the interval [start, start + duration). Final CPS equals accepted presses divided by the selected number of seconds. Mouse, touch and pen input are kept separate. Runs lose validity when the window loses focus or the tab is hidden.

Mouse and keyboard

We show observed press and release events. Untested inputs remain unknown. Key repeat does not create another initial key press. Operating-system shortcuts, browser shortcuts, input methods and some side buttons can intercept events before they reach us.

Short click intervals

The double-click check reviews adjacent primary press intervals below a selected threshold. The 80 ms starting value is exploratory, not a validated failure boundary. Increasing the threshold flags more intervals. Intentionally fast clicking can produce the same pattern.

Dragging and scrolling

The drag check separates release from cancellation and asks whether you meant to release. Pointer travel is in CSS pixels. The wheel check compares vertical event direction with your selected direction and retains pixel, line and page units separately. Event counts are not wheel notches; momentum and touchpad behavior can affect results.

Four result meanings

• Not observed: no target symptom recorded in completed runs; untested inputs remain unknown.
• Needs another check: a target observation occurred in one completed run.
• Observed in more than one run: the observation repeated under the same check settings in this session.
• Cannot judge: no valid run, interrupted input or insufficient evidence.

Next steps

Repeat the same guided steps. Then compare a fresh browser session or another browser and, if available, another device or port. Keep browser version and settings in your report. Avoid deliberately stressing painful or uncomfortable movements.

Method version

browser-v1.0 · Development baseline. Hardware reference testing and cross-browser validation remain release requirements.

Published data

Our first-hand measurements are published so they can be checked and cited, not taken on trust: see the double-click chatter reference and the machine-readable double-click input behavior dataset (versioned, with an explicit list of what has not been captured yet).