That site looks 88.46% familiar: quantifying similarity of Web page design

• Published in: Expert systems Vol. 22; no. 3; pp. 115 - 120
• Main Authors: Martine, Giselle, Rugg, Gordon
• Format: Journal Article
• Language: English
• Published: Oxford, UK Blackwell Publishing 01-07-2005; Blackwell Publishing Ltd
• Subjects: Card sorting; card sorts; Design; Expert systems; multidimensional scaling; Plagiarism; plagiarism detection; product differentiation; Research methods; Studies; Web site design; Web sites
• ISSN: 0266-4720; 1468-0394
• DOI: 10.1111/j.1468-0394.2005.00302.x

: Web page design guidelines produce a pressure towards uniformity; excessive uniformity lays a Web page designer open to accusations of plagiarism. In the past, assessment of similarity between visual products such as Web pages has involved an uncomfortably high degree of subjectivity. This paper describes a method for measuring perceived similarity of visual products which avoids previous problems with subjectivity, and which makes it possible to pool results from respondents without the need for intermediate coding. This method is based on co‐occurrence matrices derived from card sorts. It can also be applied to other areas of software development, such as systems analysis and market research.