Individual differences in verbal working memory, visuo-spatial working memory, and metacognition: learning from text in a hypertext environment



Journal Title

Journal ISSN

Volume Title


Texas Tech University


In recent years, educational institutions have increased reliance upon computerbased methods of instruction in which students learn by accessing a computer-based information system (Baird & Percival, 1999). The informatton system has been referred to as hypertext or hypermedia. The hypertext system consists of nodes. Nodes are units of information presented in text, video, audio, or pictorial formats. Links tie together the nodes, and links are used to move from one node to another (Gall & Hannafin, 1994). Prior literature suggests that the computer presentation used for hypertext may change the cognitive demands placed on the learner, because hypertext learners use links and menus to guide leaming from nodes of information (Conklin, 1987; Duchastel, 1990). The main purpose of the present study was to investigate how individual differences in verbal working memory, visuo-spatial working memory and metacognitive ability could be used to predict performance in h\pertext and linear text conditions.

The current study was designed to extend and replicate the metacognition findings of Stimson (1998), using material likely to be encountered on the Worid Wide Web. There were three sets of hypotheses encompassing the following areas: (1) differences in hypertext and linear text reading time, test scores, and perceptions about the text; (2) metacognifive predictions and judgments of performance for hypertext and linear text; and (3) differences between hypertext and linear text learning due to individual differences in verbal working memory, visuo-spatial working memory, and metacognitive ability. In general, the findings showed that visuo-spatial and metacognitive accuracy (gamma) predicted comprehension in both linear and hypertext conditions. In contrast, verbal working memory and metacognitive bias did not predict comprehension for either condition. There were no significant differences in the amount of time spend reading the nutrition information or in comprehension test scores. The participants in the linear text condition felt that the text was better organized than those in the hypertext condition. The hypotheses regarding the metacognitive predictions and judgments were not supported, although a negative correlation between bias and learning from hypertext was observed. In the hypertext condition as compared to the linear text condition, participants with less overconfidence earned higher scores on the comprehension text. That is, the correlation between hypertext and linear text comprehension test scores and bias ( r = - 0.27) was significant, as Stimson's (1998) Exp. 2 had shown.