Usability testing centers task performance rather than preference
Aliases: task usability · behavioral performance · preference performance gap
What it is
Performance-based usability testing centers success, errors, paths, time, and recovery while people perform representative tasks. Preference and satisfaction can accompany it but answer evaluative questions rather than substitute for performance. A liked interface can fail; a disliked one can still support efficient completion.
Why it happens
Performance emerges from interface cues, knowledge, goals, and context, while evaluation also reflects aesthetics, brand, and expectation. “Looks easy” skips execution demands. Completion alone also misses frustration and mistrust, so behavioral and attitudinal evidence should remain distinct and complementary.
Studying it
Prespecify success, partial success, critical error, assistance, and stopping criteria. Record action sequences and collect difficulty or confidence after tasks. Analyze behavior and ratings separately, then inspect discordant cases. Hold tasks, familiarity, and intervention protocols stable across comparisons.
Where it stops holding
Preference is the outcome when studying concept appeal, positioning, or aesthetics. Completion in collaborative or longitudinal systems may span people and weeks, beyond a short session. Success alone does not establish usability when achieved through high workload or unsafe workarounds.
Applying it
- Define observable task success and failure.
- Capture behavior before asking evaluative questions.
- Report completion, error, path, and preference separately.
- Retest changes on critical realistic tasks rather than preference scores alone.