This paper theoretically presented and empirically discussed three principles for designing useful and easy to use computer system.
This paper first presented three principles for designing useful and easy to use computer system. The authors further discussed their findings that designers often neglected or misinterpreted these principles. They also pondered why designers did so. Based on these empirical results, they re-articulated the principles, followed by a case study of applying the principles to design IBM Audio Distribution System.
- The three principles:
- Early Focus on Users and Tasks;
- Empirical Measurement;
- Iterative Design.
- Two user-oriented designs:
- Interactive Design: designers work closely with ‘a panel of expected users’;
- Participatory Design: let users participate in designing the system they are going to using.
- (What ‘Early Focus On Users’ mean is) to create a situation in which potential users can instill their knowledge and concern into the design process from the very beginning.
- Getting useful design information from prospective users is not just a matter of asking;
- Three practiced way of getting design
- Users may endorse a proposal uncritically, … They may understand the technical issues so well that they can no longer detect the difficulties in comprehension…
- (Problems of rational analysis, compared to empirical approach:)
- … it leaves things out: rational analysis does not tell you what you have to analyze;
- … systems almost always have to interact with preexisting work methods, and mismatches can be major problems.
- (Two wrong assumptions: )
- … usability work must be added to the end of the development cycle, as opposed to overlapped with it;
- … responding to tests must be time consuming.
- User testing will happen anyway: if it is not done in the developer’s lab, it will be done in the customer’s office.
- Thus, when viewed properly, a major reason for behavioral targets is that they provide a management tool to assure that system development proceeds properly.
- (What a statement of behavioral goals must contain: )
- A description of the intended users must be given, and the experimental participants to be used to represent these users in tests should be agreed upon…
- The tasks to be performed, and the circumstances in which they should be performed, must be given…
- The measurement of interest, such as learning time, errors, number of requests for help, or attitude, and the criterion values to be achieved for each, must be given…
- It appears that superior quality can be attained only when the entire user interface, including software, manuals, etc., can be designed by a single group.
Rethink the three principles:
- Early Focus on Users and Tasks
- The ultimate goal of an interface is to enable someone to do something. The design of an interactive system should (to various extent) yield the decisions to users, whether it be designers to know more about the users, or users to participate in the design process. Further, without ensuring the design is right for the users and tasks, no coding or development is really needed.
- Empirical Measurement
- Designers should also take an empirical approach to connect users and the systems, as opposed to purely rational analysis. This includes creating various media (prototype, simulations, etc.) to communicate with the users about the design; as well as evaluating users’ behavioral reactions to the designs.
- Iterative Design
- Designs and user testing should proceed iteratively one after the other. To achieve this, it is important to avoid the trap of thinking such iteration costs too much time, or is not necessary. Project teams should come up with smart methods to execute such iterative design.