A usability test asks people to do a task while you watch and record. The results are evidence, and they show where a design works and where it fails.
This lesson belongs to human computer interaction. It uses an original, fictional prototype.
How do you write an observable task?
Write a task that has a clear start and a clear end. “Reserve the book Atlas for next Tuesday” is observable. “Look around the app” is not.
Decide what you will record for each tester before the test begins.
| Measure | How to record it |
|---|---|
| Success | Did the tester finish the task? Yes or no |
| Steps | Count the taps or clicks to finish |
| Time | Seconds from start to finish |
| Trouble spots | Where the tester paused or went back |
Keep the tasks identical and the instructions the same for every tester.
Worked example: a library reservation prototype
Five students try the task “Reserve the book Atlas.” The ideal route takes 5 steps. Here are the results.
| Tester | Finished? | Steps | Seconds |
|---|---|---|---|
| 1 | Yes | 5 | 42 |
| 2 | Yes | 6 | 38 |
| 3 | No | 9 | 65 |
| 4 | Yes | 5 | 41 |
| 5 | Yes | 5 | 44 |
Completion: 4 of 5 finished, which is 4 ÷ 5 = 0.8, or four in five.
Average time of those who finished: (42 + 38 + 41 + 44) ÷ 4 = 165 ÷ 4 = 41.25 seconds.
Average steps of all five: (5 + 6 + 9 + 5 + 5) ÷ 5 = 30 ÷ 5 = 6 steps, against an ideal of 5.
What do the results say?
Tester 3 failed after 9 steps and 65 seconds. The observation note says the tester tapped “Search” three times, because the “Reserve” button sat under a menu labelled “More.”
The conclusion is specific: the Reserve button is hard to find, which is a clarity and navigation problem. The suggested fix is to show Reserve on the book’s main page.
The mistake: a conclusion with no evidence
A student writes: “The test shows the design is good.” The table shows four of five testers finished, and one failed for a clear reason. A fair conclusion reports the figures and names the problem.
Also avoid changing the task halfway, or explaining the screen to the tester during the test. Both make the results unreliable.
Check yourself
Four testers try a task. The times are 30, 36, 28 and 34 seconds, and all four finish. The steps are 4, 5, 4 and 5. What are the completion count, average time and average steps?
Answer
All four testers finished: 4 out of 4.
Average time: (30 + 36 + 28 + 34) ÷ 4 = 128 ÷ 4 = 32 seconds.
Average steps: (4 + 5 + 4 + 5) ÷ 4 = 18 ÷ 4 = 4.5 steps.
What to study next
Explain what your results mean with explaining usability decisions. For technical checks on forms, read testing interaction and data validation.
Then try the cluster practice set. If you want a teacher to review your test plan, see online one-to-one Computer Science tuition.