- Published
- Updated
Testing a device only with people close to the research team may hide difficulties faced by intended lay users. Explain the sampling approach by linking experience, language, manual skills and support needs to the actual device tasks.
Define the assessment
Do not reduce user profiles to a single characteristic such as age. Prior experience, reading habits and availability of a helper may have different effects. Collect sensitive information only when relevant to the question. State that a small usability sample does not automatically estimate population-level event frequencies.
Worked example
For example, a home measurement device is tested only with relatives of hospital employees. Task completion appears high, but those participants know clinical terminology. The team plans representation of inexperienced intended users and examines missing experience groups before generalising findings to all home users.
Practical checks
- Connect user characteristics that may affect critical tasks to the recruitment rationale.
- Record previous device training or similar-product experience and examine it alongside findings.
- Distinguish supported use from independent use when planning the participant sample.
Decision and limitations
The report should show who participated and who was not represented. Diversity alone does not demonstrate that every risk was investigated. Connect observed difficulties to tasks and design without blaming users, and base further testing on the questions that remain unresolved.
Frequently asked questions
- Is a single user characteristic sufficient?
- Usually not. Consider the experience, support and functional characteristics that may influence the tasks together.
- Can a small sample estimate population risk?
- Such testing often aims to understand problems. The design must separately support any claim about event frequency.