Running a User Day
Performing a Lite/Lab Usability Test
In contrast to task-based exploration, the Usability Test is a controlled test situation without a continuous exchange of opinions. The user is asked to solve the tasks on his or her own, working on a problem until he or she either explicitly requests assistance or obviously cannot continue without help. The Classic Usability Test also requires an implemented application, most of which can be explored with test data, but not jointly with the observer.
The advantages of the Classic Usability Test lie in the minimal influence on the user by the observer. You can see how far the users get on his or her own without any assistance to see whether or not the application is self-explanatory. Classic Usability Testing provides quantitative results that reveal information such as the solution time, the number of work steps, the number of errors and the amount of assistance given.
The developer's primary role in this phase is to observe the user. He or she should log questions that arise as well as difficulties with the application. The developers only provide assistance during the test phase if the user explicitly requests this. The user should work alone or with a partner when testing the application.
Interview Technique
There should be only one primary contact person (the observer) for each user. The role of other attendees should be explained. More than one observer per user or pair of users could bother the testers. The observer should first introduce himself and then point out that the user isn't being tested to see if he or she makes mistakes, but that the objective of the test is to detect potential weaknesses in the application in order to improve it.
The observer in the Classic Usability Test has a purely passive role and should only interfere in task processing if the user requests help.
No direct promises for concrete changes should be made and no function requirements should be discussed during the test.
If two users are solving the problem jointly, the observer should make sure that the users swap roles after each task.
Exceptions
The observer should only actively intervene while the customer is working when:
- He or she notices that the tester is losing too much time in an unplanned part of the program.
- He or she notices that the attitude of the tester threatens to become negative because of difficulties in solving the tasks.
- He or she notices that the tester cannot solve the problem without assistance.
If the user needs assistance, the observer should note what aspect of the program led to that point. Then ask the user to comment on the problem they are experiencing (for example, "Please tell me where you are stuck" or "Which function are you looking for?"), Often this results in the user solving the problem for him or herself. Obviously, the observer should never make negative comments that could disturb the user. For example, he or she should never say, "You did the wrong thing" or "You should not have used Function XYZ."
Log
In the Classic Usability Test, the observer plays a passive role and can concentrate on logging the test situation, writing down everything he or she notices. The observer should carefully record difficulties and problems requiring his or her intervention and to log the type and scope of his or her assistance.
Discuss the notes with the users after processing the individual tasks and correct them if necessary. This step should make the log more transparent for the user. If you find that you need to discuss a certain point, you can refer to the discussion in the afternoon.
Source: User Day Toolkit