Browse Books

Reviewer: Carol A. Wierzbicki

The goal of this book can be summed up in Spools statement from the preface: “This report isnt about the theory of web site design. Its based on data from our research, and describes how well (and poorly) some information-rich sites actually work when people use them to find specific answers” (p. xv). In contrast to the many highly technical Web design books available, this book is introductory, intended for those just beginning to learn about Web design dos and donts. It is a fast read, and can be read in one sitting. The conciseness and crisp layout of the book are appropriate to the subject matter. I appreciated that, in a book with so many screen shots, tables, and figures, the editors took special care to avoid breaking up the text with the figures. The study examined the usability properties of nine content-intensive popular Web sites: Cnet; Disney; Edmunds (a car and truck buyers resource); Fidelity; Hewlett Packard; Inc. (a resource for small businesses, the Web counterpart to the print magazine of the same name); the 1996 Olympics; Travelocity; and WebSaver (which provides annuity information). Two of the sites, Disney and Inc., changed their design during the course of the study. Another, the 1996 Olympics site, disappeared altogether. The authors ranked the sites according to how successful users were at finding information when asked to provide answers to four types of questions: simple fact retrieval; judgment (analyzing the information retrieved to sufficiently formulate an opinion; comparison (researching two or more facts, then comparing them to arrive at the answer); and a combined comparisonjudgment question. After the first chapter, which introduces the study, there are chapters devoted to navigation, links, within-site searching, difficulties encountered in making comparisons, readability and layout, graphic design, and user preferences. These chapters make up Part 1, “Research Results.” Part 2, “Site Scrapbook,” comprises a chapter-by-chapter examination of each sites features, outlining their strengths and weaknesses and describing problems encountered by users when searching for answers to the test questions. The screen shots included in each chapter are useful for illustrating what worked and what did not. Part 3, “Testing Sites,” describes in detail the methodology used in the testing. Users who were at least familiar with a Web browser were recruited. They were asked to find answers to the four types of questions, with each set of four tailored to the content of the site in question. The users were given two questionnaires. The first, given in mid-test, focused on how the users felt physically and mentally after searching each site. The second, given after the test, asked users to rate each site on a scale of 1 to 7 in 16 areas, such as logic of navigation, ease of finding information, quality of graphics, and the users overall productivity with the site. The book ends not with a bibliography or index, but with a kind of appendix that reads like an advertisement for the authors consulting firm. Because of the nature of this book—it is not intended as a textbook or a reference, merely as a guide for those who are thinking of testing their own Web sites—the inclusion of this information does not detract too much from its credibility. The companys Web site ( ) includes some concrete suggestions for Web design, and I would like to see a report on one of t he things UIE claims to have studied—wizards in commercial software. The firm also offers a newsletter, an email alert, and courses. Some of the conclusions drawn from the study are not surprising. For example, users find animation on a page distracting, and being sent to a different site when clicking on a link is confusing. Other conclusions—that traditional readability gauges do not apply in a Web environment, and that users often liked a site regardless of its usability—are surprising. Disney and Cnet, the sites the authors expected to do best, made a poor showing in the actual rankings, while Edmunds came in first. Several factors contribute to my impression that this book could have been more complete. First, the authors explanation of the formula they used to develop a scale for the relative rankings of each site is unclear: Answers to three of the mid-test questions…were highly correlated with successful completion of the task, users preferences, and other factors from the post-test questionnaire. We decided to use these three questions to calculate the site ratings. For each site, we took the average answer to questions 3, 4, and 6 and multiplied them together. Since each question used a scale from 1 to 7, the highest possible rating a site could achieve was 7 × 7 × 7, or 343. We then divided the sites score by 343 to place it on a scale from 1 to 100. Second, the text is seeded with numerous disclaimers (for example, “This is only a theory—we didnt talk to the site designers…” and “We cannot say that any of these sites is better or worse than the others, because we dont have information about what their developers were trying to accomplish.”) that dilute the impact of the findings. Third, I often found two of the four question types used—comparison of facts and comparison of judgment—too similar to each other to be useful in testing how users found answers to different types of questions on the Web. Fourth, Table 8.1, on page 94, uses confusing headings for the side-by-side comparison of the sites: “User Success,” “Users Like Most,” and “Users Dislike Least.” Fifth, three readability indices used—the Gunning Fog, Flesch Reading Ease, and Flesch-Kincaid Grade Level—are not adequately referenced or explained. Still, some useful conclusions about Web design can be drawn from reading this book. I now know that users would rather scroll down (to the end of a page) than drill down (through a hierarchical series of links); that designing for surfing would appear to be counter to designing for searching; and that well-differentiated (versus overly general) links have a higher likelihood of directing users to what they are looking for. This is not a book for discovering what the successful Web sites are; it is an example to follow in designing a test of ones own Web site.