This timely book explores an integrated approach to software development, and in particular the human-computer interaction component of software systems. The authors describe and propose scenario-based development (SBD), an interactive methodology that is especially useful in the early stages of analysis and development of software systems. Contrary to the traditional waterfall model of development, and in line with more recent trends in the object-oriented literature, the book proposes SBD as an iterative approach, in which due consideration is given to early user involvement. A running example (the development of an online version of a conventional science fair) is used extensively throughout the book to show in detail how the key ideas apply to each phase of development. Considerable space is devoted to the running example, with emphasis on design alternatives and tradeoffs at every stage.

Chapter 1 introduces the general concepts of usability, the software design and development process, and SBD. Chapter 2 describes requirements analysis, and introduces the virtual science fair example. Chapters 3 through 6 describe the various stages of design and development, each with extensive use of the virtual science fair example. Chapters 7 and 8 describe usability evaluation and the related topic of user documentation. Finally, chapters 9 and 10 present brief overviews of the human-computer interaction and usability engineering fields.

This book’s primary value is probably as a textbook, rather than as a reference book for working professionals. The authors provide an excellent and well-reasoned introduction to the many complexities of the field for readers who lack in-depth experience in software design and development. The running example is very useful in showing the tradeoffs one encounters at all levels. For a professional, however, the depth is a little limited, in the sense that the book stops short of describing an actual engineering discipline. From this point of view, the title is perhaps over-ambitious; the material is much closer in scope and depth to design suggestions or guidelines than to actual engineering principles. The authors themselves point out that the entire field is still at a relatively early stage of evolution, and that the portions that do have engineering strength (cognitive science models of human-computer interaction, for instance) are generally only applicable to constrained laboratory situations.

The presentation is very clear on a local level. However, the book as a whole is organized less effectively. Each chapter is designed as a relatively self-contained treatment of one stage, and as a result ends up containing a variety of material, including references and discussion of the general literature; mentions of specific new research, particularly in the areas closer to the authors’ interests; general principles and recommendations; and the running virtual science fair example. The overall reading experience is therefore marked by a certain feeling of repetitiveness and constant shifting of focus. Chapters 9 and 10 are the least effective, because their treatment is encyclopedic rather than in-depth.

Because of these considerations, I would personally describe this as a good book, rather than a “great” book. Its greatest strength lies in the lucid exposition of design tradeoffs, including the fact that they represent the norm, rather the exception, in any application design and development effort. Scenario-based development is clearly a useful technique that is well grounded in the current state of the art, and as such it should be useful to anybody who is interested in improving the usability and acceptance of software artifacts.