Review by: Dreier, Matthias (2004-07-29)

Donald A. Norman, a professor of computer science, psychology and cognitive science at Northwestern University and Stanford University, examines in this book the design of everyday things from a cognitive-psychological perspective. The book was originally published as “The Psychology of Everyday Things”. Later, the paperback issue was called “The Design of Everyday Things” – allegedly because the book was placed in the psychology sections of libraries and was therefore ignored by designers.

Norman starts the first chapter with amusing anecdotes about unusable everyday objects – doors, refrigerators, telephones, VCRs, car radios. He argues that these objects are unusable due to bad design rather than because of incompetent users. Mostly the lack of affordances is the crux of the matter. The term “affordance” refers to the perceived properties of a thing. For instance a door handle should signal whether the door has to be pushed or pulled.

In chapter two the author derives four universal design principles from the deficiencies of badly designed objects based on a psychological seven-stage model of action. Visibility: the user can tell the state of a device and possible actions by looking at it, a good conceptual model: the designer provides a coherent and consistent system image, good mappings: it is easy to determine the relationships between actions and results, and feedback: the user receives continuous feedback about the state of the system.

Chapters three and four deal with the users’ memory load and ways to reduce it: natural mappings and physical, semantic, cultural and logical constraints. An example of bad design is an array of switches that all look the same. The user has to remember the function of each switch. Of course, labelling the switches would help and is indeed very common, but if a door handle requires a label, it is most likely designed poorly.

Errors are discussed in chapter five. Norman explores the types of errors, where errors occur, and how they can be prevented. The next best thing to preventing errors is providing an undo function and reducing the severity of errors. After all, to err is human.

Chapters six and seven finally show why good design is so hard to achieve and consequently provide advice on how to design properly. Designer often have better technical skills than the users. Hence, they are not aware of possible usability flaws. Engineers are no better. They tend to add more and more features to their devices. The proposed solution is user-centred design. Features are useless as long as they are unusable. Therefore the design process needs to focus on the user, on his skills, on his intentions, on his environment. “When all else fails, standardize” is another piece of advice. Sometimes there are no natural mappings, sometimes it is difficult to visualise the current state of a device, and sometimes it is impossible to convey a conceptual model. Yes, designing everyday things is a difficult task, but that is no excuse for not trying hard.

Although the book is primarily about object design, its findings apply to websites and software interfaces as well. Norman is an exceptionally gifted story-teller. It is a pleasure to follow his thoughts. Even abstract concepts are marvellously exemplified. More advanced readers find over thirty pages of annotations, references to background literature and suggested readings. This book is a must-read for researchers and practitioners alike in the fields of usability and design.