Users need to be able to actively choose the current input or output modality

What it is

Users should be able to choose explicitly which modalities they use to input and receive feedback—touch to voice, spoken output to text, or visual alert to haptic reminder. Active choice differs from system-initiated substitution: it gives users authority over ability, privacy, setting, and the current task instead of letting the system decide which channel is “better.”

Why it happens

Modality choice depends on private, momentary conditions: occupied hands, noise, hearing or vision, a meeting, and whether a record should exist. A system can sense only part of this and may sense it wrongly; users know it directly. Explicit choice also supports exploration and learning: someone can try speech recognition, fail, and return to touch. Without an entry point, policy runs on a designer's typical situation, forcing edge tasks through unsuitable channels.

Studying it

Combine choice experiments with contextual walkthroughs. Present the same task under driving, meetings, glare, noise, occupied hands, and assistive technology, and record selections, switching time, and reselection after failure. Variables include channel combinations, consequence, privacy, and entry location; outcomes include context fit, switching cost, completion, and preference stability. Sample assistive-technology users separately because “equivalent” alternatives often carry learning costs.

Where it stops holding

Active choice does not require exposing every channel for every task. Safety-critical operations may need a fixed confirmation channel; public speech can leak private content; some devices lack hardware. Nor should choice become overhead: frequent tasks should retain a recent choice while showing it and allowing change. When the selected channel is unavailable, report why and offer a close alternative instead of silently rerouting.

Applying it

• Put modality selection near task entry and show the current input and output channels immediately.
• State each option's conditions, limits, and whether it broadcasts content, so users do not accidentally choose public audio.
• Preserve task state and unfinished input across switching; do not force a restart.
• Verification: compare an explicit selector with default policy in at least three constrained settings on completion, switching time, and perceived control.

Related

• Within the group: D5.09.2 Hidden automatic switching makes users misjudge the active modality · D5.09.3 The switch entry point should be reachable at any time, not only on particular screens
• Adjacent: D5.04.1 Noise, light, and social setting determine modality availability · D5.11.1 The system may remember modality preferences in specific contexts
• Search terms: modality choice · user control · channel preference