Usability and Evaluation
Usability and evaluation is the part of HCI concerned with defining how usable an interactive system is and with the methods, both empirical and analytical, for measuring and improving it.
Definition
Usability is the extent to which a system can be used by specified users to achieve specified goals with effectiveness, efficiency, and satisfaction in a specified context of use; evaluation is the systematic application of empirical and analytical methods to measure usability and identify problems.
Scope
This area covers the definition of usability and the methods used to assess it: empirical usability testing with users, inspection methods such as heuristic evaluation and cognitive walkthroughs, usability metrics and questionnaires for measuring effectiveness, efficiency, and satisfaction, and the cognitive models that predict interaction performance. It addresses how evaluation is planned, conducted, and interpreted. It does not cover the upstream gathering of user requirements, treated under user research methods, nor the design principles being evaluated, treated under interaction design.
Sub-topics
Core questions
- How is usability defined and broken into measurable components?
- When should evaluation use real users versus expert inspection?
- How are usability problems identified, prioritized, and fed back into design?
- How can simple cognitive models predict interaction time before testing?
Key concepts
- usability (effectiveness, efficiency, satisfaction)
- context of use
- usability testing
- heuristic evaluation
- cognitive walkthrough
- think-aloud protocol
- usability metrics
- discount usability engineering
Key theories
- Usability as effectiveness, efficiency, and satisfaction
- The ISO 9241 definition frames usability as the degree to which specified users achieve specified goals with effectiveness, efficiency, and satisfaction in a specified context, making usability a measurable, context-dependent property rather than an intrinsic feature.
- Discount usability engineering
- Nielsen argued that cheap, fast methods, small-sample user tests, heuristic evaluation, and simplified think-aloud, find most usability problems at a fraction of the cost of elaborate studies, making evaluation practical within real projects.
- Empirical and inspection methods as complements
- User testing reveals what real users actually do, while inspection methods let experts predict problems without users; combining them catches more problems than either alone.
Clinical relevance
Systematic usability evaluation is used across the software and product industries to reduce errors, support time, and abandonment; in regulated domains such as medical devices, formative and summative usability evaluation is a required part of demonstrating that a product can be used safely.
History
Usability engineering matured in the late 1980s and 1990s as software moved to mass markets. Nielsen and Molich introduced heuristic evaluation in 1990, and Nielsen's 1993 book codified discount usability methods. The ISO 9241 standards formalized the definition of usability, and evaluation became a routine phase of user-centered design.
Key figures
- Jakob Nielsen
- Rolf Molich
- Ben Shneiderman
- Clayton Lewis
Related topics
Seminal works
- nielsen1993
- nielsen1990
- iso9241
Frequently asked questions
- Is usability the same as user experience?
- No. Usability concerns how effectively, efficiently, and satisfyingly users can accomplish goals, which is measurable and largely task-focused. User experience is broader, covering the whole felt experience including emotion, aesthetics, and meaning. Good usability is usually necessary for a good experience but not sufficient on its own.
- How many users are needed for a usability test?
- For formative testing aimed at finding problems, small samples are often enough; influential work suggests a handful of users surfaces a large share of problems, so iterating with several small tests is usually more valuable than one large study. Summative tests that estimate metrics precisely need larger, carefully chosen samples.