CONDUCTING A TEST #6

Establishing tasks and scenarios.

Establishing tasks: The tasks must be real, which means they are representative of what users will do with the product when released. Naturally, a test cannot cover all possible tasks a user may do with a product. Therefore tasks must be selected carefully in order to maximize their effectiveness (represent as many of the uses as possible) and make use of precious test time. Some criteria that may be used to select tasks are (as suggested by Dumas & Redish, 1993): those that probe potential usability problems; tasks suggested from test teams' concerns and experience, and; tasks that users will do with the product.

Establishing task scenarios Once the tasks have been selected the scenarios in which they will be presented must be created. A scenario makes the group of tasks to be completed more realistic and creates an environment that forces a user in a specific direction. You can then observe whether the user figures out how to complete the task and whether the method they use matches designers' expectations/intent. It is important to remember that the scenario should be as short as possible, use the users' terminology, and gives the user enough information to do the task(s). Additionally, the tasks and scenarios need not be matched one to one. A scenarios can consist of many tasks or just one, whatever is most efficient and matches the task flow of the predicted user. Another important point to remember when designing scenarios is that the order of tasks as presented to the participant should start with simpler tasks and build to more complex ones. This approach will allow the participant to gain confidence as they go through the tasks and will not frustrate participants on the first task. This could affect their attitude, thus performance, for the rest of the test (Dumas and Redish, 1993) .