ABSTRACT
This paper explores visual attributes that define web page visual appeal. A cluster analysis and subsequent regression analysis were undertaken to assess if ratings of appeal could be predicted from a particular combination of five visual attributes. Three samples of web pages used in different studies all seeking to understand how quickly visual appeal judgments are formed were used to support our argument that it is not only necessary to understand what makes web pages appealing but that it is equally important to understand what distinguishes those from unappealing web pages. We investigated if the attributes used to identify the highly appealing pages could also identify those rated lowest in appeal. That was supported by the results reported.
