MSc-IT Study Material
June 2010 Edition

Computer Science Department, University of Cape Town
| MIT Notes Home | Edition Home |

Guidelines

A guideline is a rule about designing interactive systems. They range from very general (‘Keep the user informed about what the user is doing’ from Nielsen 1993) to the very specific (‘Use white space between long groups of controls on menus or in short groups when screen real estate is not an issue’ from the Open Look style guide).

Shneiderman’s ‘Designing the user interface’ is to a large extent a collection of design guidelines, but explanations are included to show why those guidelines are the way they are. (For example look at box 7.2 on page 264 which gives a list of form filling guidelines and then the text below which explains the reason for those guidelines.)

There are two components to a guideline: the guideline itself and the rational behind it. The ultimate idea being that if the guidelines are sound and unambiguous enough then they could be presented to designers without the rationale behind them. The designers could then apply the guidelines, almost automatically, and be sure of coming up with an usable system.

Such an approach has precedence in all sort of situations; we use sets of rules and guidelines without needing to know the theory that underlies them frequently. A cookbook describes how to make an omelette and if its rules are followed then a perfectly good omelette will be arrived at. The cook does not need to know the science of what happens when the proteins in the egg white are heated in order to successfully cook an omelette.

Rationales for guidelines come from different places and can therefore give the guidelines different levels of authority. Much academic study has attempted to link guidelines to sound scientific theory and if such a link can be established then the guideline will have a considerable level of authority. Other sets of guidelines are based around common sense thinking or possibility previous experience; a practitioner will write down what worked for his design, possibly with some explanation of why.

It is important to distinguish between ‘standards’ which we discussed in the previous unit and the guidelines we are discussing here. They are both rules about what designers should or should not do, but standards carry much more authority; they have to be obeyed. Guidelines are much looser and are often transgressed. Because of the higher authority of standards then they must have a well defined and inspectable rationale behind them.

Problems with guidelines

Unfortunately the science behind usability is not sound enough that this sort of reliance on guidelines can be usefully achieved. The designer will in many cases need to understand, to some extent, the science or rationale underlying the guidelines so that they can be intelligently applied. In many cases usability guidelines contradict one another, so it is important to understand their rationale in order to decide which is the most appropriate.