Applied user performance modeling in industry - A case study from medical imaging

Abstract

To determine ways of improving a user interface - so that a person's task produces fewer errors or takes less time - is a major goal of user interface designers. Usability testing is an established and proven method for revealing these user performance metrics, but is rather time-consuming, resource intensive and requires at least functional prototypes. Therefore, it may not always be the optimal choice during the development of very complex, expensive and context-specific applications like those in medical imaging. A promising alternative is to simulate user performance with computational models based on psychological models. In this paper the Performance Modeling inventory (PMI), which was developed based on Card, Moran and Newell's (1980) Keystroke Level Model for estimating user performance with medical imaging systems, is introduced for the first time. © Springer-Verlag Berlin Heidelberg 2007.