The body of positive research, hypotheses, and comment concerning graphical systems is being challenged by some studies, findings, and opinions that indicate that graphical representation and interaction may not necessarily always be better. Indeed, in some cases, it may be poorer than pure textual or alphanumeric displays. Trying to force all system components into a graphical format may be doing a disservice to the user. Some also feel that, as graphical systems become increasingly sophisticated and continue to expand, interfaces become increasingly more complex, sometimes arcane, and even bizarre. Among the disadvantages put forth are the following:
Greater design complexity.
The elements and techniques available to the graphical screen designer far outnumber those that were at the disposal of the text-based screen designer. Controls and basic alternatives must be chosen from a pile of choices numbering in excess of 50. (Conversely, alternatives available to the text-based screen designer numbered about 15.) This design potential may not necessarily result in better design, unless the choices are thoughtfully selected and consistently and simply applied. Proper window types must also be chosen and colours selected from a seemingly unending rainbow of alternatives. With graphics, the skill of the designer is increasingly challenged. Poor design can undermine acceptance.
Learning still necessary.
The first time one encounters many graphical systems, what to do is not immediately obvious. The user may not know the meanings of many words and icons. It is often not possible to guess their meanings, especially the more arbitrary symbols. The user may also have to learn how to use a pointing device. A severe learning and remembering requirement is imposed on many users, and it takes a while to get up to speed. A text-based system could easily be structured to incorporate a set of clear instructions: (1) Do this, (2) now do this, and so on.
System providers estimate that becoming accustomed to a graphical interface should require about eight hours of training. Other experts say the learning time is closer to 20 or 30 hours.
Lack of experimentally-derived design guidelines.
The graphical interface is still burdened today by a lack of widely available experimentally-derived design guidelines. Early on, more developer interest existed in solving technical rather than usability issues, so few studies to aid in making design decisions were performed. Today studies being performed in usability laboratories are rarely published. This occurs because of several factors. First, builders of platforms and packages will not publish their study results because they want to maintain a competitive advantage. If they find a better way to do something, or present something, why tell the competition? Let them make the same mistake, or find the answer themselves.
Second, the studies are often specific to a particular function or task. They may not be generally applicable. Third, it takes time and effort to publish something. The developer in today’s office seldom has the time. Finally, it is also difficult to develop studies evaluating design alternatives because of increased GUI complexity. Too many variables that must be controlled make meaningful cause and-effect relationships very difficult to uncover.
Consequently, there is too little understanding of how most design aspects relate to productivity and satisfaction.
Inconsistencies in technique and terminology.
Many differences in technique, terminology, and look and feel exist among various graphical system providers, and even among successive versions of the same system. These inconsistencies occur because of copyright and legal implications, product differentiation considerations, and our expanding knowledge of the interface. The result is that learning, and relearning, for both designers and users is much more difficult than it should be.
Working domain is the present.
While direct-manipulation systems provide context, they also require the user to work in the present. Hulteen (1988), in a takeoff on WYSIWYG, suggests “What you see is all you get.” Walker (1989) argued that language takes you out of the here and now and the visually present. Language, she continues, makes it easier to find things.
Not always familiar. Symbolic representations may not be as familiar as words ornumbers. People have been exposed to words and numbers for a long time. Research has found that numeric symbols elicit faster responses than graphic symbols in a visual search task. One developer had to modify a new system during testing by replacing iconic representations with a textual outline format. The users, lawyers, were unfamiliar with icons and demanded a more familiar format.
Human comprehension limitations.
Human limitations may also exist in terms of one’s ability to deal with the increased complexity of the graphical interface. The variety of visual displays can still challenge all but the most sophisticated users. The number of different icons that can be introduced is also restricted because of limitations in human comprehension. Studies continually find that the number of different symbols a person can differentiate and deal with is much more limited than text. Some researchers note that claims for the easy understanding of pictograms are exaggerated, and that recognizing icons requires much perceptual learning, abstracting ability, and intelligence.
The motor skills required may also challenge all but the most sophisticated users. Correctly double-clicking a mouse, for example, is difficult for some people.
Window manipulation requirements. Window handling and manipulation time are still excessive and repetitive. This wastes time and interrupts the decision making needed to perform tasks and jobs.
Production limitations.
The number of symbols that can be clearly produced using today’s technology is still limited. A body of recognizable symbols must be produced that are equally legible and equally recognizable using differing technologies. This is extremely difficult today.
Few tested icons exist.
Icons, like typefaces, must appear in different sizes, weights, and styles. As with text, an entire font of clearly recognizable symbols must be developed. It is not a question of simply developing an icon and enlarging or reducing it. Changing an icon’s size can differentially affect symbol line widths, open areas, and so forth, dramatically affecting its recognisability. Typeface design is literally the product of 300 years of experimentation and study. Icons must be researched, designed, tested, and then introduced into the marketplace. The consequences of poor or improper design will be confusion and lower productivity for users.
Inefficient for touch typists.
For an experienced touch typist, the keyboard is a very fast and powerful device. Moving a mouse or some other pointing mechanism may be slower.
Inefficient for expert users.
Inefficiencies develop when there are more objects and actions than can fit on the screen. Concatenation for a command language is impossible.
Not always the preferred style of interaction.
Not all users prefer a pure iconic interface. A study comparing commands illustrated by icons, icons with text, or text-only, found that users preferred alternatives with textual captions.
Not always fastest style of interaction.
Another study has found that graphic instructions on an automated bank teller machine were inferior to textual instructions.
Increased chances of clutter and confusion.
A graphical system does not guarantee elimination of clutter on a screen. Instead, the chance for clutter is increased, thereby increasing the possibility of confusion. How much screen clutter one can deal with is open to speculation. The possibility that clutter may exist is evidenced by the fact that many people, when working with a window, expand it to fill the entire display screen. This may be done to reduce visual screen clutter. Mori and Hayashi (1993) found that visible windows, not the focus of attention, degraded performance in the window being worked on.
The futz and fiddle factor.
With the proliferation of computer games, computer usage can be waste of time. Stromoski (1993) estimates that five hours a week in the office are spent playing and tinkering. Experts have said that the most used program in Microsoft Windows is Solitaire! Tinkering includes activities such as creating garish documents reflecting almost every object property (font size, style, colour, and so on) available.
Futzing and fiddling does have some benefits, however. It is a tool for learning how to use a mouse, for example, and it is a vehicle for exploring the system and becoming familiar with its capabilities. It is of value when done in moderation.
May consume more screen space.
Not all applications will consume less screen space. A listing of names and telephone numbers in a textual format will be more efficient to scan than a card file.
Hardware limitations.
Good design also requires hardware of adequate power, processing speed, screen resolution, and graphic capability. Insufficiencies in these areas can prevent a graphic system’s full potential from being realized.
Atin Dasgupta is director and founder of Leveljam. Web Design Company Mumbai Leveljam is an internet services agency from Mumbai, India focussed on providing cutting edge design and business solutions to the services & manufacturing sector with innovative approaches and advanced methodologies. Our Services include Web Design & Branding, Logo Design, Flash Animation, Application Development, Website Development, E-commerce Development, Content Management Solutions, Mobile Applications Development, Search Engine Optimization, Internet Marketing Solutions, Social Media Visibility & Management, Technical Support and Customer Service. For more information please visit our website www.leveljam.com
