Attention is split by the environment; focused use cannot be assumed
Aliases: situationally-induced impairment · divided attention · SIID · mobile distraction
What it is
Someone is still in the interface, but eyes and hands are also given to the pavement, a conversation, a bag. When the surroundings take attention, capability drops for a while: situational impairment (also situationally-induced impairments and disabilities). The task is not paused and the session need not be short; what fails is the assumption of someone seated, both eyes on the glass. Scanning a code at a gate, changing volume in a kitchen, confirming in the rain—failure comes from occupied channels, not from a spent time budget, and not from the app being backgrounded. This entry is only about divided attention and situational load. It is not about how short a launch is, and not about whether state survives a cut.
Why it happens
Walking, talking, and carrying are secondary tasks that compete with the interface for eyes, ears, hands, and cognitive slots. Vision is owed to obstacles, hearing to a greeting, one hand to a rail; the remaining thumb is unsteady. The person is not refusing to read. Continuous gaze fragments into short glances, so fine dragging and long paragraphs become expensive. Sun, gloves, and vibration stack on top: hits miss, words are misread. Unlike interruption, the person never left the task; they continued it on degraded perception–motor bandwidth. An interface still laid out for laboratory focus—tiny legal lines, adjacent dangerous buttons, sliders that must be seen to the end—concentrates errors when environmental load is high. People slow down, bring the phone closer, stop walking to tap. Those compensations are evidence that load has already exceeded walk-and-use capacity.
Studying it
Use dual-task: walk at a set pace or listen to speech that must be paraphrased, while doing the interface task. The site can be a lab corridor, a walkway with obstacles, or a real street with safety controls. Field observation counts how often people stop walking in order to operate.
Independent variables: walking speed or secondary-task load, one hand versus two, noise and light, target size and copy length. Dependent variables: time and errors, missed reading, gait change, number of stops, subjective load.
Lab secondary tasks are a steady rhythm; street load is bursty. An interface that passes a corridor test can still fail at a junction. Do not extrapolate seated usability scores to mobile situations. Eye tracking is hard while walking; stops and task failures are sturdier field measures. Report “completed while walking” separately from “had to stand still.”
Where it stops holding
Seated, lying down, or with the phone on a table, situational impairment is weak and desktop-like density and precision can hold. Driving is a different safety regime; walk-and-glance conclusions do not move into the car. Eyes-free, speech-only tasks contest the visual channel differently. Highly practiced users can finish in shorter glances, but a dangerous confirmation must not rely on practice to skip a look. A cradle that holds both hands still and the screen fixed is outside this entry.
Applying it
- Make critical confirms short sentences and large targets; do not bury irreversible actions in copy that needs continuous reading.
- Avoid gestures that succeed only if a drag hits an exact end; prefer a tap over a thin slider.
- Put feedback where residual vision can catch it: enough contrast, a state change large enough not to be a grey numeral shift.
- Verify by having people complete the main path one-handed while walking (in a safe space). Record stops, misses, and unread text, then run the same path seated in quiet. Failures that appear only while walking are load, not an unusable control.