To explore an application together with a user, the application development must have reached a stage where substantial parts of the application are ready to run using test data.
The advantage of a joint task-based exploration over the Classic Usability Test is that you can discuss problems with the user directly in the context of the current task. This gives authentic and accurate feedback and comments about the usability of the application.
In contrast to paper prototypes, you cannot make spontaneous design changes here. You can only record suggestions, not implement them. In contrast to Classic Usability Testing, you have to dispense with a controlled test situation in favor of a more intensive and continuous exchange of opinions.
In task-based exploration, prepared tasks are used as a guide for exploring an application.
A good overall scenario is important and helpful because exploration is more natural if the individual tasks fit together.
In contrast to the Classic Usability Test, you can leave the normal task path at some points to explore other aspects of the application, but make sure that processing the tasks does not end up as free exploring. If it does, point out that free exploration will be possible at the end.
After processing the tasks, and perhaps after processing each area of the application, ask the user to try out the application again in his or her preferred way or how he or she would normally use it.
When jointly exploring an application, the interviewer must take advantage of the fact that he or she is participating directly in processing the tasks and can therefore directly address problems as they occur. Under no circumstances, however, should he or she take the lead, suggest solutions or explain the application. He or she is thus a passive partner with regard to processing the tasks, but an active partner regarding problems that occur.
In addition to directly discussing the problems that occur, the interviewer should observe the tester closely and notice any non-verbal reactions.
The interviewer should immediately follow up on all suspicions about the tester's satisfaction and all interpretations about user problems. The interviewer has a unique opportunity to have his or her suspicions confirmed or rejected immediately by the user. The quality of the user feedback is always better the closer you are to a real working situation. This is also true for probing spontaneous design ideas resulting from the observations. Instead of keeping design ideas and interpretations in your mind, you should communicate them immediately to confirm them in the context of usage. The graphic below illustrates this principle of checking ones own interpretations and ideas with the user during the observation.
Write down as much as possible. This is true for the testing sequences and for all important statements made during the conversation. The log should contain not only your interpretation or summary, but also the original quotes made by the tester.
Source: User Day Toolkit