A Formal Method for Designing the Interface of Interactive Systems

This research investigates the use of a formal method as a tool for designing and evaluating the interlace of interactive computer systems. The method enables one to make a formal description of the user actions and system responses, to identify flaws and inconsisten cies of the system, to make predictions about the user difficulties and errors with various parts of the interface, to test these predictions, and modify the interface according to the predictions that are confirmed. The new interlace version that comes out is expected to be more consistent and predictable. and therefore, easier to learn and use. It can be improved iteraiively. An experiment was run to test our predictions about subjects' errors and difficulties with a graphics system, light subjects with knowledge about computers participated in the study. In a first session, they were asked to learn how to use the system. In a second session (two days later), they had to complete a questionnaire about this system. Results confirmed our predictions (with very few exceptions). Several recommendations are proposed for improving the user interface. Despite a few shortcomings, this formal method greatly helps in testing and improving the consistency and predictability of a system.