Using Eye Movements to Measure Intrinsic, Extraneous, and Germane Load in a Multimedia Learning Environment

• Published in: Journal of educational psychology 2020-10, Vol.112 (7), p.1338-1352
• Main Authors: Zu, Tianlong, Hutson, John, Loschky, Lester C., Rebello, N. Sanjay
• Format: Article
• Language: English
• Subjects: Attention Control; Cognitive Processes; Computer Assisted Instruction; Difficulty Level; Eye Movements; Female; Human; Human Channel Capacity; Learning; Learning Environment; Learning Processes; Male; Measurement; Measurement Techniques; Multimedia; Multimedia Instruction; Reaction Time; Replication (Evaluation); Short Term Memory; Undergraduate Students; Visual Measures
• ISSN: 0022-0663; 1939-2176
• DOI: 10.1037/edu0000441

In a previous study, DeLeeuw and Mayer (2008) found support for the triarchic model of cognitive load (Sweller, Van Merriënboer, & Paas, 1998, 2019) by showing that three different metrics could be used to independently measure 3 hypothesized types of cognitive load: intrinsic, extraneous, and germane. However, 2 of the 3 metrics that the authors used were intrusive in nature because learning had to be stopped momentarily to complete the measures. The current study extends the design of DeLeeuw and Mayer (2008) by investigating whether learners' eye movement behavior can be used to measure the three proposed types of cognitive load without interrupting learning. During a 1-hr experiment, we presented a multimedia lesson explaining the mechanism of electric motors to participants who had low prior knowledge of this topic. First, we replicated the main results of DeLeeuw and Mayer (2008), providing further support for the triarchic structure of cognitive load. Second, we identified eye movement measures that differentiated the three types of cognitive load. These findings were independent of participants' working memory capacity. Together, these results provide further evidence for the triarchic nature of cognitive load (Sweller et al., 1998, 2019), and are a first step toward online measures of cognitive load that could potentially be implemented into computer assisted learning technologies. Educational Impact and Implications Statement Sweller has proposed that during learning activities, students can experience three types of cognitive load: (a) intrinsic load, depends upon the inherent complexity of the learning material, and the learner's prior knowledge; (b) extraneous load, depends upon instructional design, is due to cognitive process that do not contribute to schema construction or automatization, and should be minimized; and (c) germane load, is due to cognitive processes that contribute to schema construction and automatization, and should be maximized. We replicated the main results of a seminal study that differentiated between the three types of cognitive load using a specific set of manipulations. Further, our study showed that different eye movement-based metrics were sensitive to different cognitive load subtypes. This provides further evidence demonstrating the validity of the triarchic model of cognitive load theory. This study also demonstrates that one can measure Sweller's three different cognitive load types using physiological measure