Issue
Article
Vol.30 No.3, July 1998
Article
Issue
Issue |
Article |
Vol.30 No.3, July 1998 |
Article |
Issue |
A cursory visit to the computing section of your local bookshop will reveal an odd omission in the literature of human-computer interaction. We are well supplied with texts discussing the "front end" activities of requirements gathering, such as task analysis, use cases, scenarios, and a variety of other techniques for discovering and describing user requirements. Equally, there is a plenitude of books on the "back end" activities, such as the design and layout of screens and controls for a variety of implementation environments. (Indeed, it is hard to believe that the world needs quite so many books on Web site design; on the other hand, the evidence of the Web itself suggests that few of these books are in the hands of Web site designers.) There are also many well-defined processes for evaluating the usability of an interface once it has been designed, even when it is just a low-fidelity paper prototype. One step in the design process, however, has been oddly neglected: the step that transforms user requirements into an initial interface design (which can then be refined through established techniques). This is the "gap" of this book's subtitle.
For a variety of reasons, this transformation from requirements to design has traditionally been regarded by many involved in user interface design as the most difficult to do and the hardest to define. This is at least partially due to the vast number of ways one might implement a given set of requirements. This process may devolve to relying on the creative skills of the local guru, or open brainstorming amongst a team; or, as Larry Wood, the editor, describes it in the introductory chapter, "then a little magic happens". Of course, as Larry Wood goes on to point out, this process is no more magical than any other creative design process. This book aims to make explicit the conscious processes of experienced designers, and thereby enable other designers to exploit, repeat, and perhaps even improve a variety of methods that have been successfully field-tested.
Larry Wood has been interested in this gap for a number of years, and this interest led him, along with Ron Zeno, to organize a workshop on the topic at CHI 96 in Vancouver. The workshop attracted a broad range of participants from North America and Europe; from industry and academia; and from hardware, software, and specialist consulting organizations. The workshop in turn led to the idea of the book. The text comprises contributions from many of the participants, each chapter addressing the gap from a particular individual's or group's perspective.
Chapter 2, a Bellcore team effort, is paradoxically both the most practical and the most frustrating chapter. The authors have clearly defined a design process (coincidentally called "The Bridge") and environment that will produce successful designs. This process can readily be replicated in other projects and organizations. Even though a book's worth of material has been abridged to a single chapter, the reader still gets a good feeling for the process. I was frustrated, however, that the authors did not justify more explicitly one important assumption integral to their method, namely that the resulting interface will be centered on the objects in the user's model, rather than on the processes that the user carries out. Although they acknowledge that such object oriented designs are not the best choice when a task may only be done in one way, I was not convinced that highly restricted processes are as rare as the authors suggest. (For a contrasting viewpoint, see chapter 8.)
In contrast to the breadth of chapter 2, chapter 3 (by Thomas Graefe of Digital) focuses specifically on the nature of the relationship between requirements and design. Graefe considers the gap as a transformation between a represented world (requirements) and a representing world. This transformation is mediated by conceptual models and metaphors, and also by the "grammar" of the representing world -- the syntax, style, and conventions of the specific implementation environment (e.g., Windows 95). By making explicit the conceptual model and metaphors, the designer helps to ensure that they are applied consistently, coherently, and completely. Graefe also provides some interesting observations on why designers cannot rely on their own intuition to guide design decisions, and provides some techniques for provoking the creative process. Chapter 4, by Frank Ludolph (Sun Microsystems), expands on this idea to present a series of models and transformations between them. Ludolph also describes some useful design techniques including the task tree, which is a functional breakdown of user tasks. Some readers might find these two chapters slightly abstract; nonetheless they contain highly useful perspectives on the transformation process.
Lightweight, relatively informal techniques intended for smaller scale designs are described in chapter 5 by Andrew Monk (University of York, United Kingdom). Monk suggests that the gap can be crossed most successfully if the designer first minimizes the gap, so that designs can readily be shown to correspond to requirements. One of the techniques described here is the Work Objective Description, or WOD, as a representation of user tasks. At first sight the WOD is simply an inversion of the hierarchical task analysis; however, by putting the goal explicitly at the top-most level and then breaking it down into successive levels of sub-tasks, Monk suggests that the technique encourages creative or even radical approaches to carrying out the task. The hierarchical task analysis by contrast, says Monk, tends merely to document existing processes. Monk also emphasizes the importance of dialog design as the first level of implementation design.
Chapter 6 provides an interesting change of viewpoint. In this chapter, the authors (from Linné Data) use a first person narrative to provide a fascinating introspection on the designer's experience. In doing so they make explicit some of the creativity involved in bridging the gap, such as generating and reflecting on ideas. They describe "Bubbling", an associative clustering technique related to clustering techniques used in other creative activities such as writing, once again emphasizing the creativity at the heart of UI design. This focus on creativity is reiterated in chapter 9, in which Jean Scholtz (National Institute of Standards and Technology) and Tony Salvador (Intel Corporation) address the idea of "Systematic Creativity". Systematic creativity is a framework for inclusive design particularly aimed at brand new products driven by new technologies. This theme recurs in chapter 11 where Colin Smith (Nortel Technology) deals with the introduction of "new-generation products". Smith emphasizes the importance, in this environment, of establishing high-level user goals through exploratory design.
In stark contrast, chapter 8 by Sabine Rohlfs (IF Interface Consulting) presents the process of redesigning complex legacy systems. Rohlfs discusses the resource limitations often associated with legacy work, and emphasizes the need to consider carefully current tasks and their malfunctions. In this context Rohlfs maintains that a redesigned legacy system often lends itself most naturally to a task-oriented, rather than object-oriented, dialog style.
In chapter 7, Martin Rantzer (Ericsson Radio Systems) describes the Delta method. Rantzer emphasizes the importance of considering user documentation as an integral part of the usability of the design. Rantzer also discusses the importance of the Design Room, a sanctuary where design activities can be conducted relatively free of interruption. In chapter 10, Kevin Simpson (Financial Models Company) presents a related idea, the UI War Room. Simpson also discusses a technique for managing multiple perspectives and ultimately, synthesizing prototypes. Highly refreshing in this chapter is Simpson's acknowledgment of the fact that these techniques do not make design a simple process; it still requires a skillful designer to work the `magic'.
In summary, this collation of real-world experiences certainly has much to offer to everyone who is involved in making the `magic' happen. However, this is not a book to be absorbed at a single sitting. The variety of writing styles may disorientate the reader, and the huge selection of design techniques offered might leave the reader slightly overwhelmed. There is much value in this book, and it deserves being digested a chapter at a time. This book not only bridges the gap between requirements and design, it also bridges an important gap on our bookshelves.
User Interface Design: Bridging the Gap
Larry E. Wood, Editor
Publication date: 1998
ISBN: 0-8493-3125-0
Published by: CRC Press, Incorporated
Carl Zetie
czetie@us.oracle.com
Carl Zetie is currently Manager of Product Direction for the Developer/2000 products at Oracle Corporation. He has worked with a variety of development tools for the past 12 years, and became interested in usability about five years ago. He is the author of Practical User Interface Design (McGraw-Hill, 1995).
Issue |
Article |
Vol.30 No.3, July 1998 |
Article |
Issue |