A Guide to GOMS Model Usability Evaluation using GOMSL and GLEAN4

The standard accepted technique for developing a usable system, empirical user testing, is based on iterative testing and design revision using actual users to test the system and help identify usability problems. It is widely agreed that this approach, inherited from Human Factors, does indeed work when carefully applied (Landauer, 1995). However, Card, Moran, & Newell (1983) have argued, and many HCI researchers have agreed (e.g., see Butler, Bennett, Polson, & Karat, 1989), that empirical user testing is too slow and expensive for modern software development practice, especially when difficult-to-get domain experts are the target user group. One response has been the development of "discount" or "inspection" methods for assessing the usability of an interface design quickly and at low cost (Nielsen & Mack, 1994). However, another response, which has been evolving since the seminal Card, Moran, and Newell work, is the concept of engineering models for usability. Analogously to the models used in other engineering disciplines, engineering models for usability produce quantitative predictions of how well humans will be able to perform tasks with a proposed design. Such predictions can be used as a surrogate for actual empirical user data, making it possible to iterate through design revisions and evaluations much more rapidly. Furthermore, unlike purely empirical assessments, an engineering model for an interface design can capture the essence of the design in an inspectable representation, making it easier to reuse successful design insights in the future.