From users’ task knowledge to high‐level interface specification
Suzanne Sebillotte, Dominique L. Scapin · International Journal of Human-Computer Interaction · 1994
In the effort to define appropriate design and evaluation methods for interface usability, research work is being carried out to study the matching between users’ task characteristics and interface characteristics. This article proposes a methodology: following the task description (using Methode An‐alytique de Description; MAD), select task‐based data for defining informational constraints for the interface. The air traffic control task was described by applying MAD. Even though it was found to be suitable for capturing task‐based knowledge, the method showed the need for further refinements. This led to a more systematic way of selecting human factors rules as candidates for generic inference rules. These rules consist of premises describing specific configurations of users’ tasks (e.g., a particular informational need, a relationship between objectives, etc.) and conclusions that indicate specific requirements for the interface to conform to (e.g., a particular information display, a simultaneous access to two related functions, etc.). The air traffic control task description and the first interface specification attempts, even though they were exploratory, demonstrated the feasibility of the approach—that is, of generating interface requirements from task configurations, through human factors rules.