After its first appearance in the user-interface of the Xerox 'STAR' office workstation, icons have been gaining popularity. Many studies have been performed in relation to the design of appropriate icons. However, most of these research concentrated on 'how' to represent the icons to make them look good. Very few studies investigated 'what' should be included in an icon to make it work better....[ Read more ]

After its first appearance in the user-interface of the Xerox 'STAR' office workstation, icons have been gaining popularity. Many studies have been performed in relation to the design of appropriate icons. However, most of these research concentrated on 'how' to represent the icons to make them look good. Very few studies investigated 'what' should be included in an icon to make it work better.

This study was aimed to investigate 'what' should be represented in an icon to make it effective. Three factors affecting icon effectiveness were hypothesized: dominance, uniqueness, and ambiguity. Dominance is a factor intrinsic to a function. An element (verb or object) of a function is said to be dominant if the other element can be inferred from the first one. Uniqueness is related to the quality of design. If an icon has perceptual immediacy, then it can be said to be unique. If multiple meanings are possible in a single context by one user, the icon can be said to be ambiguous.

Two experiments were conducted to investigate how the three hypothesized factors affect user performance in terms of time and accuracy. The results of the experiments showed that dominance and uniqueness are two main factors affecting user's response time in an iconic interface, whereas ambiguity is the main source of accuracy problems. A performance model showing the relationships among these factors and user performance were suggested based on the findings of the study. Finally, some useful recommendations for icon designers to design effective icons and further research directions are provided.