Mobile contexts usually lack a stable support surface

What it is

When a phone or handheld device is used while walking, standing, or gripped in one hand, there is usually no external fixed surface — a desk, an armrest — available. It's not just that the finger performing the pointing motion has no support; the hand holding the device itself has no anchor either. This is fundamentally different from a desktop scenario, where the device itself sits stably on the desk and, at most, only the operating hand lacks support — in a mobile context, even the frame of reference (the device) is itself shaking.

Why it happens

The hand holding the phone has to keep it aloft using its own muscles, so that hand carries the same physiological tremor as any unsupported hand, plus body sway from walking — and the screen itself (the reference plane the other hand is aiming at) moves along with it. Walking or unsteady standing adds a further low-frequency whole-body oscillation on top. The finger performing the tap has to contend with two sources of motion at once: its own jitter, and the drift of the target plane itself. The combined effective error is substantially larger than the finger jitter alone seen at a desk.

Studying it

A common approach collects touchscreen tap data in real mobile contexts — walking, riding transit, standing at a stop — and compares the resulting landing points against seated, stationary baseline conditions, quantifying the added positional error and error-rate increase from being in motion. Some studies separately measure the device's own acceleration/angular velocity while in motion to quantify how much the reference plane itself drifts.

Methodological note: precision loss measured while participants "simulate walking" in a lab (e.g., on a treadmill) tends to underestimate the loss seen on a real street under real levels of distraction — genuine mobile use adds visual distraction and obstacle avoidance that lab conditions don't include, so lab data is a lower bound, not an upper one.

Where it stops holding

This applies to mobile contexts with no additional support available; if the user happens to be leaning against a wall, holding a railing, or has set the device on a bag or table, this constraint no longer holds and the situation is closer to supported, stationary operation. Older adults or people using mobility aids are more likely to have both hands occupied simultaneously, making the constraint more severe than in typical mobile use.

Applying it

• Interfaces intended for mobile use should default to sizing touch targets and spacing for unsupported conditions, rather than reusing precision data measured on desktop or tablet.
• Critical actions (payment confirmation, sending) should avoid requiring a single precise hit on a small target while the user is walking — provide a secondary confirmation or a more generous tolerance zone rather than assuming the reference frame is stationary.
• To verify: have the same users complete an identical tap task both seated and while walking, and compare the error rate and the spread of landing points — a large gap indicates the current target size only holds up in stationary conditions.

Related

• Same group: A8.14.1 Wrist or pinky support markedly improves pointing precision · A8.14.2 Unsupported operation reduces precision and fatigues faster · A8.14.3 A support surface constrains the available range of motion
• Nearby: A8.15 Physiological tremor · A8.18 Symmetric bimanual coordination · B1.02 Deviations from Fitts' Law on touchscreens
• Search terms: on-the-go input · mobile touch accuracy · walking while typing · situational impairment