Multimedia learning is defined as building mental representations from materials that involve both verbal (spoken or written text) and pictorial information (static or dynamic visualizations; Mayer, 2005). Many studies on the effectiveness of multimedia learning have been conducted, often inspired by Mayer’s cognitive theory of multimedia learning (see Mayer, 2005) and Sweller’s cognitive load theory (see Sweller, 2005). However, these studies have mainly drawn conclusions about the cognitive effects of different types of multimedia learning materials based on (transfer test) performance measures, and measures of cognitive load and time-on-task, without directly investigating the processes underlying these effects. Hence, the empirical work presented in this double symposium focuses on detailed analyses of the processes underlying the learning effects of different types of multimedia materials by means of eye tracking. Because eye movement data can provide detailed insight into the allocation of (visual) attention and processing demands, eye tracking is a valuable tool for such studies –albeit one that is little used in educational research. In this double symposium, studies are presented that focus on learning from a variety of multimedia materials that include dynamic visualizations, static visualizations, written text, and narrated text, in varying compositions.

Although multiple external representations can have benefits, especially for learning complex and new ideas, they are often not as effective as expected.. The present study employs eye tracking methodology to take a closer look at how learners use different external representations in learning from worked examples, how these activities are related to learning outcomes, and how well intended cognitive functions of multiple representations match to the functions as perceived by the learners. 16 (predominantly psychology) students studied worked examples on the application of probability principles, each consisting of a text (the problem formulation), an equation (representing the solution), and a tree diagram (that was intended to mediate between the concrete text and the highly abstract equation). During the learning phase the gazes of the participant were recorded. After the learning phase, the participants saw a gaze replay of their viewing behavior and were asked to think-aloud. The distribution of fixation durations on different representations indicates that single representations were not neglected. Rather, the participants switched frequently between different external representations which might indicate that they are not processed independently from one another. Yet, transitions between representations were not per se beneficial. For example, transitions between diagrams and equations were valuable only for learners with better learning prerequisites. For learners with poorer learning prerequisites, frequent transition obviously indicated mapping difficulties. For all learners, frequent transitions between equation and text were dysfunctional, pointing to the mediating function of the diagrams. Hence, the function of transitions seems to depend on characteristics of the learners as well as on characteristics of the representations involved. Finally, the verbal protocols revealed a mismatch between intended and perceived functions. The results suggests to inform learners more fully of the intended functions and/or to make the intended functions more salient.