PDF of this tutorialEarly design decisions are usually those concerning the product concept, its architecture, and its priorities. As a result, these are the decisions that are the most costly to change later. If these decisions are not user-centered, the end product will not be usable. For example, adding well-designed icons to a flawed menu structure will not rescue a poor product.
Figure 2. Involve Users Early and Continuously
For this reason, users should be involved as early as possible in the design process. Users usually contribute to early efforts to gather information through observation, questionnaires, focus groups, interviews, or more detailed task analysis. At this stage, designers will build models of the users' domain and establish task priorities and relationships. As was noted, users may not be good at articulating their requirements or describing or predicting their own behavior. For this reason, field observation of user behavior is often most effective.
Users are also better at critiquing an existing HMI than designing one from scratch. However, user involvement must be cost-effective. Simply placing users in front of a new application and asking for improvements will lead to an unprioritized wish list. This is why prototyping is crucial. Users should be given a number of alternative designs—whether high-level or detailed—to compare and critique. The alternatives will help them generate more ideas and also show that their comments are welcome and useful.
As the design becomes relatively stable, user activities are aimed at refining and validating detailed design. Usability testing becomes most effective when measuring performance and productivity—including error rates and causes—and validating terminology and icons. Typically, users will be asked to carry out specific tasks designed to test parts of the interface or to address particular design issues that could not be resolved earlier. Well-designed user trials will get maximum results for the time and effort invested by designers and users.
