from Evaluation of Human-Computer Interaction Whilst Using the World Wide Web
by Nicolas J. Blaza · The Buckinghamshire College, a college of Brunel University · May 1996
3.1 Introduction
User interfaces are changing as companies re-examine the whole concept of human-computer interaction (Seymour, 1995). Software designers of graphical user interfaces have a wide range of hardware and software tools available to them. For example, they may be limited to a static graphics display, or have the computing power to have fully rendered animation.
3.2 Case Study: Designing the Xerox Star User Interface
[The original section described the Xerox Star user interface case study. Key passages elsewhere in the report (see §2.7) summarise the same material: Xerox formulated the user's conceptual model before software was written, spent thirty person-years on the interface design, and pioneered concepts such as bit-mapped displays, icons, the desktop metaphor and WYSIWYG editing.]
3.3 Principles Used
The following main goals were pursued in designing the Star interface:
- Familiar user's conceptual model
- Seeing and pointing versus remembering and typing
- What you see is what you get (WYSIWYG)
- Universal commands
- Consistency
- Simplicity
- User tailorability
A user's conceptual model is, according to Preece and Keller (1990), "a set of concepts a person gradually acquires to explain the behaviour of a system. It is the model developed in the mind of the user that enables that person to undertake and interact with the system." For example, the "desktop" is a metaphor — the work space the user can work on the screen; and the use of icons and windows shows the principle of seeing and pointing versus remembering and typing in operation. A page on the screen was the same as a printed page, coining the phrase WYSIWYG. Commands such as 'MOVE', 'COPY' and 'DELETE' were used as universal commands, constant throughout the system.
A well-designed system makes everything relevant to a task visible on the screen. It does not hide things under CODE+key combinations or force the user to remember conventions that may be obscure or arbitrary.
3.4 The Apple GUI
The Graphical User Interface of the Apple Macintosh computer builds on the ideas put forward in the Star project and was one of the earliest attempts to use a visual language to communicate instructions to the user. The interface is the 'place' where the computer and the user meet. In order to improve communication, Apple used the metaphor of the desktop, building on the user's direct experience of the office by using icons, or images, to represent folders, files, and dustbins.
The Macintosh interface was unique, and certainly the most mature expression of a developing language within computing that later formed the basis of many similar products. Other computer software producers began to recognise the value of this predominately visual environment and designed their interfaces accordingly. The Microsoft Corporation launched a software product, Windows, mainly for the IBM personal computer and its clones. A software application called X Windows was also developed for the expanding workstation market.
Of all the other ways of interacting with the computer, none has been as widely accepted as the mouse. Users took to this strange, single-handed device with real swiftness. The mouse was an integral part of the total WIMP (Windows, Icons, Mouse and Pointers) environment that characterised much of personal computing in the 1980s (Baker, 1993).
3.5 Icons
Icons are an integral part of any GUI; some applications, such as the Xerox Star and Apple Macintosh systems, for example, use icons to represent the "electronic desktop". Preece and Keller (1990) list the possible reasons for using icons in human-computer interfaces:
- They are visually more distinctive than a set of words.
- They can represent a lot more information in a small space (space being at a premium on a VDU screen).
- They may be subjectively more desirable than text.
In the view of van Dam (1984): "It has been found that an interface based on menus and icons is preferred by most people over a strictly alphanumeric interface because, when these graphical features are properly designed, they seem more natural, are easier to learn and use, require little memorising and result in fewer mistakes." The graphical design of Web browsers has had this effect, as the results in chapter 6 show.
In chapter 5 the author highlights the prominent use of icons when navigating around the Web. Preece and Keller (1990) offer the following constructive advice on designing icons; the author has illustrated this advice with examples from Web browsers.
- Icons are best used for concrete concepts and are most effective as a miniature representation of the physical object to which they refer (Preece & Keller, 1990). In Netscape, for example, the sketch of a home is used to indicate the user can go to their home page at the click of an icon (figure 3.1).
Figure 3.1 — Netscape's "home" icon representing the home page of the user.
- The more abstract the concept, the harder it is to find an acceptable icon. A good example of this is the Mosaic icon for adding a file to the user's HotList (figure 3.2). To a novice user this icon could mean something completely different to its actual function.
Figure 3.2 — Mosaic "burning page" icon representing adding the current file to the user's "hot list".
- Provide icon text labels for novice users and infrequent tasks, which can be turned off as required. Alternatively, show an "icon key" on their screen.
- For common actions provide a command equivalent, e.g. to go to the Bookmark option in Netscape the user can either go to the drop-down menu under Bookmarks or type Control+B. Having to type Control X, Y or Z would not follow as consistently in a user's familiar conceptual model.
© 1996 Nicolas J. Blaza. Reconstructed online edition, 2026.
Best viewed in Netscape Navigator 2.0 or higher at 800×600 resolution.
