The XAOS metric - Understanding visual complexity as measure of usability

Abstract

The visual complexity of an interface is a crucial factor for usability, since it influences the cognitive load and forms expectations about the subjacent software or system. In this paper we propose a novel method that uses entropy, structure and functions, to calculate the visual complexity of a website. Our method is evaluated against a well known approach of using the file size of color jpeg images for determining visual complexity. Both methods were applied on a dataset consisting of images of 30 different websites. These websites were also evaluated with a web survey. We found a strong correlation for both methods on subjective ratings of visual complexity and structure. This suggests both methods to be reliable for determination of visual complexity. © 2010 Springer-Verlag.