Choosing your own adventure: hyperlinks and their effects on memory

Abstract

For this experiment, 39 students from a large Midwestern university were exposed to a total of ten articles, two main articles about 150 words in length, called parent page articles, and eight linked articles. Each of the articles contained four embedded links each. Two of the linked articles contained related content, while the other two contained unrelated content. Readers were then tested on their cued recall and recognition of the articles to gauge how the embedded linked articles affected those two aspects of memory. Lang's Limited Capacity Model for Mediated Message Processing (LC4MP) was used a basis for the hypotheses. This experiment also made hyperlink cuts analogous to television cuts as in Lang, Geiger, Strickwerda and Sumner (1993). Based on those two factors, the hypotheses predicted that readers would better remember related information. The results found were counterintuitive. Readers showed better overall recognition and recall for parent pages when compared to linked pages. They also had better recall of unrelated linked information when compared to related linked information and when compared to both related and linked information.