Book title: Guidelines for Screen Design

Blackwell Scientific Publications, 1990, ISBN 0-632-02686-3.

Andrew Dillon

This item is not the definitive copy. Please use the following citation when referencing this material: Dillon, A., (1990) A Review of Rivlin et al - Screen Design Guidelines. Hypermedia, 2(2), 171-173.

This slim volume has emerged, according to the preface, from a workshop on screen design held as part of the ESRC-funded Information Technology in Education Research Programme (InTER). At this workshop, fifteen experts from various fields such as art and design, human-computer interaction and education divided into small teams to focus their discussion on issues such as design planning, layout, graphics and so forth. The results were obviously productive enough to convince the participants that there was a book in there somewhere and hey presto, another set of guidelines emerges.

If size is anything to go by then the reader might initially conclude that the various authors do not have a lot to say. Excluding front matter and references, there are only 80 pages to this book and liberal use of white space within it adds to a sense of the lightweight. However, the book has a comfortable feeling about it, is easy to handle and its contents, devoid of superfluous ramble, are surprisingly seductive. If you have ever used guideline texts before, particularly the kind with numerous cryptic cross-references that flout their own recommendations on ease of use, you’ll be pleasantly surprised with this one.

The book is interesting in that it takes a wide view of the relevant issues in screen design and places them in the general context of a design process. It is divided into five chapters covering communication (i.e., what is the software about), layout (i.e., organising the display), text (i.e., size and arrangement), graphics (i.e., types and their functions) and interaction (i.e., the user interface). Each is prefaced with an overview and list of contents, and finished with a set of summary points. Within chapters the material is organised such that guidelines are presented on the left hand page with their underlying rationale outlined opposite. The beauty of this approach is that one is never left asking why such a guideline is advocated or what supporting evidence exists. You may disagree with a guideline but you are presented with the means to find out more and thereby deduce why the authors proposed it (although critical appraisal of this is another matter).

Another positive attribute of the book is that it briefly explains the importance of understanding the limitations of the underlying hardware and software of an application, reminding designers that it is all too easy to devise ideas that are too costly or even impossible to implement with current technology. It conveys these ideas in simple terms and suggests seeking specialist help if in doubt.

The problem I have with this book is the one I have with most sets of guidelines and is, ultimately, fatal –some of the recommendations are invalid. Guidelines are best employed as an aide memoire by those who are familiar with the field or as a first port of call for those wishing to learn. Any inaccuracies can then be tempered by experience or further learning. Where this is not made clear, problems abound and with this set I can see such problems Usually, my disagreements with their conclusions reflect our different interpretations of the literature and the emphases placed on certain variables rather than the belief that the authors have necessarily got it wrong. For example, on the issue of controlling the update of text on screen the guideline here is most definitely “use paging rather than scrolling” which is qualified with a rationale statement that although scrolling is probably easy to use it may be inappropriate in some situations and the rate of scrolling will never suit all users. I, on the other hand, would contend that scrolling tends to be preferred by regular users, variance in preferred rate is not much of a problem, and that anyway, the issue ultimately depends on the task and text under consideration. For lengthy texts, simple paging would be a nightmare for non-linear access.

The thorny issue of colour however, is where I believe that the authors are totally wrong. While they correctly state that colour should be used sparingly and the text/background contrast should be high, their list of recommended combinations includes red on green and red on cyan, despite the facts that the former renders reading impossible for those with certain forms colour blindness as well as being universally warned against for use with VDUs on the basis of empirical evidence (e.g., Pace 1984) and the latter is virtually unreadable in the colour plates they use as examples. The use of black on white is cautioned against despite recent evidence suggesting it is optimal for reading text on screen (see e.g., Gould et al 1987). Their most recent suggested reference on colour is a paper from 1982.

The problem with this is that one poor guideline can ruin an otherwise useful tool. Reading their recommendations on colour startled me into thinking that if they got this wrong, why should I believe any of it? This would be unfortunate as the text is generally good. I am tempted to believe that the shortcomings stem from their over-willingness to apply findings from the paper medium to the electronic one, which we should all know by now is fraught with difficulties.

To get the maximum benefit from this book I would recommend that it is read from start to finish before use. In this way the reader can gain the user-oriented perspective that is so important for the sensible application of guidelines. Dipping in and pulling out ‘rules’ such as “paging is better than scrolling” or “avoid right justification” is acceptable if one appreciates the ifs and buts of the situation. Extracting the guideline “red on green is a good colour combination” in the absence of any consideration that this might be slightly unergonomic is not. The latter is the type of usage that gives guidelines a bad name and results in horrific user interfaces developed in slavish accordance to so called “proven” (read “published”) design principles. Thank God there isn’t an index!

The editors state that they have tried to avoid being directive, no doubt aware of the general problem just identified. This may have been their aim but if any design principle holds true it is that there is no legislating for users and what they try to do. That goes for books of guidelines as much as for software. You have been warned.

References

Gould, J.D., Alfaro, L., Finn, R., Haupt, B. and Minuto, A. (1987) Reading from CRT displays can be as fast as reading from paper. Human Factors 26 (5), 497-517.

Pace B. (1984) Color combinations and contrast reversals in VDUs. Proc. of the Human Factors Society 28th Annual Meeting. Human Factors Society: Santa Monica 326-330.