Using Multiple Ways to Investigate Cognitive Load Theory in the Context of Physics Instruction

Cognitive load theory (CLT) (Sweller 1988, 1998, 2010) provides us a guiding framework for designing instructional materials. CLT differentiates three subtypes of cognitive load: intrinsic, extraneous, and germane cognitive load. The three cognitive loads are theorized based on the number of simultaneously processed elements in working memory. Intrinsic cognitive load depends upon the number of interacting elements in the instructional material that are related to the learning objective. Extraneous cognitive load is the mental resources allocated to processing unnecessary information which does not contribute to learning as caused by non- optimal instructional procedure. It is determined by the number of interacting elements which are not related to learning goal. Both intrinsic and extraneous load vary according to prior knowledge of learners. Germane cognitive load is indirectly related to interacting elements. It represents the cognitive resources deployed for processing intrinsic load, chunking information and constructing and automating schema. Germane cognitive load is related to level of motivation of the learner. Given this triarchic model of cognitive load and their different roles in learning activities, different learning outcomes can be expected depending upon the characteristics of the educational materials, learner characteristics, and instructional setting. In three experiments, we investigated cognitive load theory following different approaches. Given the triarchic nature of cognitive load construct, it is critical to find non- intrusive ways to measure cognitive load. In study one, we replicated and extended a previous landmark study to investigate the use of eye movements related metrics to measure the three kinds of cognitive load independently. We also collected working memory capacity of students using a cognitive operation-span task. Two of the three types of cognitive load (intrinsic and extraneous) were directly manipulated, and the third type of cognitive load (germane) was indirectly ascertained. We found that different eye-movement based parameters were most sensitive to different types of cognitive load. These results indicate that it is possible to monitor the three kinds of cognitive load separately using eye movement parameters. We also compared the up-to-date cognitive load theory model with an alternative model using a multi-level model analysis and we found that Sweller’s (2010) up-to-date model is supported by our data. In educational settings, active learning based methodologies such as peer instruction have been shown to be effective in facilitating students’ conceptual understanding. In study two, we discussed the effect of peer interaction on conceptual test performance of students from a cognitive load perspective. Based on the literature, a self-reported cognitive load survey was developed to measure each type of cognitive load. We found that a certain level of prior knowledge is necessary for peer interaction to work and that peer interaction is effective mainly through significantly decreasing the intrinsic load experienced by students, even though it may increase the extraneous load. In study three, we compared the effect of guided instruction in the form of worked examples using narrated-animated video solutions and semi-guided instruction using visual cues on students’ performance, shift of visual attention during transfer, and extraneous cognitive load during learning. We found that multimedia video solutions can be more effective in promoting transfer performance of learners than visual cues. We also found evidence that guided instruction in the form of multimedia video solutions can decrease extraneous cognitive load of students during learning, more so than semi-guided instruction using visual cues