A good usability test is usually structured as a series of discrete tasks assigned to a participant. Each task should explore a small specific function or piece of content on the site. (This can come from any task analysis you performed earlier in your design cycle.)
Once you have chosen what to test, how do you begin to design tasks?
Example: Your site includes a basic book search and an advanced book search. The best approach would be to create two sequential tasks. The first would be to search for a book by title. The second would be to perform a more complex query requiring the advanced search.
Dividing a usability test into a series of short tasks has several advantages:
- It makes your instructions to the participant simpler. They have to remember less at each step.
- If the participant encounters problems early in one long complex task, you won't be able to get results for later stages of the task. Shorter tasks reduce this risk.
- Following and analyzing a participant's actions is simpler when they are performing a series of short tasks. They are less likely to get off track, and when they do, it is easier to get them back on track.