Motor impairments raise accidental activation well above typical rates

What it is

Give the same set of targets to people with impaired motor control—tremor, spasticity, limited range, low strength, ataxia, or a bouncing vehicle and gloves—and the rate of hitting a non-target rises well above the control group. That is not one diagnosis. Once motor noise exceeds target spacing, discrete clicks become a high-error channel. An interface that treats every click as intent writes that noise down as a command.

Why it happens

Pointing is modelled as a clean landing: down, up, that rectangle. Motor impairment pours spatial and temporal jitter into the channel. Landings spread, contact breaks and resumes, one intended tap becomes two events. As targets pack tighter, the landing cloud covers a neighbour faster—a geometric fact, not a particular physiological curve.

Different conditions share one interface outcome: a wrong event reaches the app. Spasticity may overshoot; limited reach may plant the whole hand on the next key; a moving screen is itself the jitter. The accidental-activation layer has to handle “wrong activations happen, and more often,” not to classify which tremor is on the hand.

Studying it

Pay participants with limited motor control and a control group. On identical sizes and spacing, run everyday clicks (list rows, adjacent icons, a tab bar). Log wrong-target hits, repeated triggers, and whether the person later says “that was not the tap I meant.”

Independent variables: size and spacing, whether a destructive neighbour is flush, input channel (touch, mouse, switch). Dependent variables: wrong-target rate, repeat-activation rate, task-level failures (entered the wrong flow).

Control-group success does not transfer. “Community-dwelling mild” samples undercount clinic populations. Tremor simulators are a preview, not a substitute.

Where it stops holding

Cognitive errors (misread label) and motor misses must be logged apart: copy versus tolerance. Huge isolated targets drop the rate and weaken the claim. Switch scanning errors are early or late beats, not spatial scatter; do not compare them to touch wrong-target rates. Blaming every miss on small targets will miss cases where the target is large but a single activation commits.

Applying it

• Move high-consequence controls off the high-frequency hit band; do not pack delete and send with everyday actions.
• Leave a gap beside adjacent icons; visual alignment is not a reason to glue a destructive action to the primary button.
• Verify: ask someone with unstable motor control to invoke a secondary action on a named row, and count non-target activations. If controls almost never miss and the target user repeatedly hits the neighbour, the rate gap already stands.

Related

• Same group: J3.09.2 Destructive actions that complete on a single activation carry the highest risk · J3.09.3 Undo fits this population better than confirmation dialogs
• Nearby: A8.16 Pathological and Intention Tremor · J3.04 Target Size Minimum · A11.03.7 Handling hand tremor with touch debounce and mis-tap filtering
• Search terms: accidental activation · motor impairment · mis-tap rate