Description
Select, write, time and order usability test tasks so they probe likely problems and matter to users, phrased as realistic goals without interface words or hints.
Testing samples a few tasks from many, so task choice determines what can be found. Selection draws on likely usability problems (the first criterion, by analogy with software testing that assumes bugs exist), designers' concerns, tasks hard to recover from, and new, critical or frequent tasks; an importance-times-doubt matrix (each rated 1 to 5) picks about five feature clusters for a 60 to 90 minute session. Each task becomes a short scenario in users' words, stating the goal and needed data but not the steps, with a clear end state. Early, open-ended tasks let users bring real needs; later, specific tasks with an evaluator-only solution key probe a clarified design, optionally with one unsolvable task. Time is estimated as expert time multiplied by 3 for simple tasks up to 10 for complex ones. Tasks are ordered in a natural sequence, easiest first, and piloted.
Attributes
- Discipline
- Usability & Evaluation
- Loss type
- Active
- Goal
- Evaluate usability, Prioritise
- Stage
- Evaluation
- Data
- Analytical / expert
- Effort
- Days
Steps
- List candidate tasks from likely problems, designers' concerns, hard-to-recover tasks and new, critical or frequent user tasks.
- Rate candidate features 1 to 5 for importance and for design doubt, multiply, and keep the top scorers (about five clusters for a 60 to 90 minute session), each with a one-sentence question.
- Write each task as a short scenario in users' language, with a lifelike motive, the needed data and a clear end state, avoiding interface terms and hints.
- Use open-ended tasks when needs are unclear and specific tasks with an evaluator-only solution key when probing a clarified design; consider one unsolvable task.
- Estimate time per task as expert time x 3 (simple) to x 10 (complex or data entry), sizing to roughly 5 to 10 tasks per hour.
- List resources, seeded data and reset procedures per task, and order tasks in natural sequence, easiest and most important early.
- Pilot the tasks, fix ambiguity, and present them in writing one at a time.
Sources
- A Practical Guide to Usability Testing (Dumas & Redish, 1999)
- Usability Testing: A Practitioner's Guide (Hertzum, 2020)
- Observing the User Experience (Goodman et al., 2012)
- Benchmarking the User Experience (Sauro, 2018)
- Handbook of Usability Testing (Rubin & Chisnell, 2008)
- Usability Engineering (Nielsen, 1993)
- Customer Analytics for Dummies (Sauro, 2015)
Also known as
- Task selection and ordering for usability tests
- Selecting tasks to test
- Task scenario writing
- Usability test scenarios
- Open-ended and specific test tasks
- Solution key
- Importance x doubt matrix
- Task-time estimation
Categories
- User Research
- Usability & Evaluation