Towards Metric-based Usability Evaluation of Online Web Interfaces

Abstract

Current approaches to web interface evaluation are tedious or do not provide sufficient information. Thus, we propose a new metric-based method building on interaction data and usability models. This would enable internet companies to evaluate interfaces at faster iteration cycles but poses new requirements to usability instruments. As a first step, we present INUIT an instrument aiming at this specific purpose. A confirmatory factor analysis showed that INUIT can reasonably well describe real-world perceptions of usability while being compatible with the desired metric-based approach.