Journal Issue: Children and Electronic Media Volume 18 Number 1 Spring 2008
Learning from Hypertext
Hypertexts—dynamic texts, such as a website or multimedia software program, presented on a computer in a nonlinear fashion—offer a number of advantageous possibilities for learning. Hypertexts are interactive, allowing users to take in information at their own pace in the way they are most likely to derive meaning from it.97 Hypertexts are open-ended; they allow readers to choose the information they want to retrieve and the order in which they want to retrieve it.98 In fact, readers build their own text as they navigate through the information presented.99 Typically, hypertexts recruit and sustain high levels of attention.100
With hypertexts, readers must create the structure of the text based on their own knowledge, whereas in traditional texts, readers use the existing structure of the text to make inferences that enhance comprehension.101 Hypertexts thus require additional cognitive skills, as readers are responsible for determining what information they need to further increase their understanding of the topic and how to access it.102 Research has focused on comprehension and control of hypertext.
Several studies have assessed learning from hypertexts.103 In a review of all quantitative studies of hypermedia and learning outcomes published between 1990 and 1996, Andrew Dillon and Ralph Gabbard found no overall comprehension advantages for hypermedia (even across a variety of comprehension measures) over paper presentations.104 However, hypermedia did offer significant advantages for particular tasks, such as visual categorization and discrimination and searches through large amounts of information.105
Readers' prior knowledge of a topic likely affects their comprehension of hypertexts. In one study with adults, prior knowledge improved recall from the text and also influenced how users navigated through the reading environment.106 Readers lacking prior knowledge may have difficulty navigating the hypertext, as they may find it hard to find the information they need.
Interest in content has been associated with easier, more efficient navigation through the text, whereas interest in dynamic text features, such as sound effects and video, has been associated with less comprehension.107
Increased control may offer advantages for some hypertext users. However, the benefits of increased control may vary with the ability of the user. Complexity may, in fact, hinder performance in students by confusing them.108 Some studies report a user preference for hypertexts offering control, even though learning may not be improved.109
Almost all studies of hypertext navigation have focused on adults. Kimberly Lawless and several colleagues, however, studied children's navigational strategies through hypertext. Fourth-, fifth-, and sixth-grade children completed a domain knowledge pre-assessment, individual and situational interest pre-surveys, and post-tests of recall. In addition, the computer recorded the path navigated by each user. Based on the data, the study identified distinct navigational profiles, similar to those for adults. Most students, the “knowledge seekers,” focused on the information portions of the hypertext. A smaller group of students, the “feature seekers,” spent most of their time exploring features, such as animation and movies. A third group of students, “apathetic hypertext users,” spent little overall time with the hypertext. The most knowledgeable students were more likely to be the apathetic users; the least knowledgeable, the feature seekers. The knowledge seekers fell in between. The authors concluded that prior knowledge affects navigational strategy, in that it may enhance interest in content.110
Research on learning from hypertext is limited, especially with regard to children. Dillon and Gabbard point out that the research suffers from a host of methodological flaws, limiting the conclusions that can be drawn. They argue for greater focus on the design variables responsible for different learning outcomes, as well as how those design variables interact with individual differences in users.111