Do dependable systems need good user interfaces?

Abstract

More than in other classes of systems, trust in user interfaces (UI) needs to be a reality in dependable systems. In particular, for a system that needs dependability, a poorly implemented UI can mislead the user and make her/him discredit the system. Trust in computational systems involves technical aspects, such as computational infrastructure, storage space, and service composition, but also aspects of human-computer interaction. While technical aspects have been largely studied, there are few studies about human-computer interaction regarding trust, mainly when dependable computing is the target. The perception of trust as a user’s feeling that leads him to advance in the use of the system passes first of all through the user interface, even more so when dependability is his concern. In this work, we present the results of experiments with fifty users on a quality model to assess the trustworthiness of computational systems focused on user interface. The goal is to complement the trustworthiness measurement of the whole system with interface trustworthiness measurement, besides considering dependability, security, connectivity, scalability, and isolation, among others. By using 3 online bank websites, which need to be dependable due to financial loss, it was possible to observe that the approach is feasible to be applied to dependable online systems. The major aim is to help users know if the system they are using is trustworthy.