Guidelines for Authoring Comprehensible Web Pages and Evaluating Their Success

In recent years, numerous articles, books, and even Web sites have offered advice for designing hypertext documents, multimedia, and Web sites. This advice is often based on common sense and professional experience, not research, though the same thing can be said about any type of instruction. But the bigger problem is that these guides spend only a few pages, if any, on actual writing topics—for example, content selection and organization, paragraph structure, word usage, tone, and so on. To write comprehensible Web pages or to evaluate the comprehensibility of Web pages, writers and evaluators often rely on guidelines for writing and evaluating print documents. This default approach is based on the assumption that readers comprehend information in the same way from print and online documents, perhaps a sound assumption and perhaps not. Research studies have analyzed the effect of the reading medium (paper versus screen) on many user outcome variables: comprehension, recall, speed and accuracy of search and retrieval, learning, task speed, reading strategy, attitude, motivation, and fatigue (for example, Gambrell and others 1987; Egan and others 1989; Foltz 1996; Aust and others 1993; Yang 1991–1992; Gijar and others 1998; excellent review articles have been written by van der Geest [1994] and Dillon [1992].) Whether medium has an effect on reader outcomes depends on the tasks and variables that researchers assess, as well as the equipment and materials used, but the results for comprehension and recall are equivocal: some studies find better comprehension and recall with print documents, some find the opposite, and some find no difference. Many researchers in the 1980s claimed that print documents and hypertext significantly differed in terms of their structure, maintaining that print documents are linear while hypertext documents are not. Recently, however, there has been a discussion to the effect that linear text is not always linear and that the two mediums may be more similar than one might think. In print documents, readers jump around, looking at tables of contents, indexes, figures, tables, appendixes, footnotes, and glossaries. Reading of print materials can in fact be more nonlinear than reading of hypertext because the routes with print documents are endless and the routes in hypertext are constrained by links (Foltz 1996). Rouet and Levonen (1996) point out that perhaps printed text and hypertext do not differ greatly in terms of their linearity but that hypertext allows one to more easily take advantage of nonlinear features. The ease may involve the click of a mouse, but the very act of making and following navigation decisions in hypertext imposes an additional cognitive load on the reader. Once users of a Web page locate the content they are seeking by scanning the page, they become real “readers” of the Web page, and like readers of print materials, they try to make sense of sentences, paragraphs, and pages. In fact, many components of the reading process should be quite similar for readers of print and online text. A brief review of how readers approach text and comprehend it will prove useful in understanding how Web pages can best serve their readers. (Readers desiring more detailed information than can be found here may find Britton and Graesser’s [1996] Models of understanding text of interest.)