Double-tap and long-press succeed less often when motor control is limited

What it is

A compound pointer gesture adds a time or count constraint on top of one aim: two taps inside a window, or contact held inside tolerance for a duration. When motor control is limited, both succeed less often than a single tap. They are not the same act. They share one statistical fact: stacking a time parameter on pointing raises failure on its own.

Why it happens

A single tap asks whether the landing is in the rectangle. A double tap also asks spacing and delay between two landings—the second comes too soon, too late (two singles), or drifts to a neighbour. A long press also asks whether contact can stay unbroken and inside for the whole wait. Both stretch an already hard hit across time, so error integrates.

Failure need not share one physiology. Jitter wrecks the second tap and the still hold; low strength makes people lift early; slow judgment makes the hold feel “long enough” before the clock agrees. The cross-gesture claim is that extra time constraint, not how one gesture’s progress bar should look. If a compound gesture is still the only way to a feature, missed activation and failed activation share a door: some people never get in, and those who do are more likely to enter the wrong room.

Studying it

On the same target size, compare three primitives: single tap, double tap, long press. Motor-limited versus control. Log success, and misclassification (double tap as two singles, long press as a drag).

Independent variables: double-tap window, hold duration, target size. Dependent variables: success per primitive, cross-gesture misclassification.

The task should be “invoke the second meaning on a named row”; neighbours turn drift into the wrong action. A full-screen button undercounts compound-gesture failure.

Where it stops holding

When switch, voice, or a hardware key fires the same feature, the compound pointer primitive is not used and the barrier moves. System “touch accommodations” can widen the window, slow everyone, and still not stop drift. Children who will not wait and older or tremor users who cannot hold still are different mechanisms; do not pool the evidence. Platform-specific long-press diagnosis (menu, preview, drag start) belongs to the touch-primitive layer; here only the cross-gesture success contrast is in scope.

Applying it

• Do not make a required feature reachable only by double tap or long press; offer a single-activation path to the same result.
• Put second meanings on a visible button or menu, not on a clever gesture as the only channel.
• Verify: ask someone with an unstable hand to double-tap and long-press a named target ten times each. If controls pass and the target user mostly produces two singles or early lifts, the compound gesture is not a viable sole channel.

Related

• Same group: J3.13.1 Fire on release, not press, so the pointer can still cancel · J3.13.2 Buttons passed during a scroll or drag must not count as clicks · J3.13.4 After a miss-tap, recovery should be one undo, not repeating the whole flow
• Nearby: C3.03.5 Long-press is a barrier for motor-impaired and older users · J3.05 Single Pointer Alternatives · J3.11 Path-based Gesture Alternatives
• Search terms: double tap · long press · compound pointer gesture