Website Consistency

Website Consistency

by Dr. Bob Bailey

June, 2001

How consistent do web pages need to be in order to not degrade the performance of experienced users?

Ant Ozok and Gavriel Salvendy at Purdue University studied this question with some surprising results. They believed that participants in their study would perform better and be more satisfied using Web pages that had a consistent rather than inconsistent interface design.

Two of the earliest studies on computer interface consistency reported that tasks performed using more consistent interfaces resulted in

a reduction of task completion times,
a reduction in errors, and
an increase in user satisfaction

(Schneider and Shiffrin, 1977; Schneider, Dumais and Shiffrin, 1984). A more recent study (Eberts, 1997) found that learning time decreased as consistency increased.

Adamson and Wallace (1997) suggested three components of consistency:

physical (visual characteristics, including colors, labels, locations, shapes and sizes, etc.),
communicational (methods of moving within and between pages, within menus, etc.), and
conceptual (metaphor, language, stereotypes, etc.).

Ozok and Salvendy created four versions of the same set of web pages. One was reported to be "totally consistent" while each of the others was modified in order to be physically, communicationally or conceptually inconsistent.

Experienced web users, with an average age of 26 years, performed a series of tasks using the four different websites. The investigators found that only the website with the "physically" inconsistent interface degraded performance. It did so by increasing the number of errors made by users. None of the inconsistent interfaces reliably slowed users or reduced their satisfaction with any of the websites (when compared with the "totally consistent" website).

Adamson and Wallace (1997) also had found that "physically" inconsistent interfaces increased the error rate in graphical user interfaces. Even though one or two previous studies had found that "communication" inconsistencies resulted in degraded performance and satisfaction (not found in the Ozok and Salvendy study), no studies have shown that "conceptual" inconsistency affects either performance or satisfaction.

It appears that the lack of consistency in websites may not degrade human performance or user satisfaction of experienced users as much as some people think that it does. Also, when there is a problem, it is much more likely to be an increase in errors rather than slowing of user performance or lack of satisfaction with the website.

References

Adamson, P.J. and Wallace, F.L., A comparison between consistent and inconsistent graphical user interfaces, Technical Report, Department of Computer and Information Sciences, University of Northern Florida, Jacksonville (1997).

Eberts, R.E., Cognitive modeling, In G. Salvendy (ed.), Handbook of Human Factors and Ergonomics - Second Edition, New York: John Wiley & Sons (1997).

Grudin, J., The case against user interface consistency, Communications of the ACM, 32, 1164-1173 (1989).

Ozok, A. A. and Salvendy, G., Measuring consistency of Web page design and its effects on performance and satisfaction, Ergonomics, Vol. 43, No. 4, 443-460 (2000).

Schneider, W., Dumais, S.T. and Shiffrin, R.M., Automatic and control processing and attention, Varieties of Attention, New York: Academic Press, 1-27 (1984).

Schneider, W. and Shiffrin, R.M., Controlled and automatic human information processing: detection, search and attention, Psychological Review, 84, 1-66 (1997).