The effects of visual representation fidelity and self-explanation prompts on mental model formation and cognitive load in adaptive learning environment

Abstract

The purpose of the study was to investigate the effects of different types of visual representation fidelity (progression visual fidelity and no progression visual fidelity) and self-explanation prompts (adaptive fading self-explanation prompts and fixed self explanation prompts) on mental model formation (mental model by diagram test, general knowledge, and knowledge inference) and cognitive load (extraneous load, intrinsic load, and germane load) as well as the relationship between mental model and cognitive load. The instructional methods that are often assumed to be effective for novice learners would also be effective for more experienced learners, but at the same time, these instructional methods can lose their effectiveness or even have adverse effects for more proficient learners (expertise reversal effect). A potential method for preventing the expertise reversal effect would be to design an adaptive learning that allows learners to engage in personalized learning experiences by dynamically tailoring the level of instructional support according to the corresponding expertise level of the individual learner. Based on this theoretical background, the purpose of this study was to design visual representation fidelity and self-explanation prompts by applying an adaptive approach. In addition, it was expected that progression visual fidelity combined with adaptive self explanation prompts would provide the most significant benefit to mental model and cognitive load. For this purpose, the asynchronous e-learning environment was developed Unity software. A total of 69 undergraduate students from a university in Seoul, Korea participated in this study. The students were randomly assigned to one of four treatment groups when they logged into the asynchronous e-learning environment. Of them, data from all 62 students was used for a series of MANOVA and ANOVA. According to these findings, learners who had learned under the progression visual fidelity conditions obtained significantly higher scores than those who had no progression visual fidelity on the mental model by diagram test. In other words, those who received the support of the increasing visual fidelity condition improved their accuracy from pre-test to post-test when given the task of drawing the human heart, compared to those who received the no progression visual fidelity condition. In contrast to the mental model by diagram test, learners who were provided no progression visual fidelity gained significantly more general knowledge about the human heart and circulatory system than learners who were provided with progression visual fidelity. Finally, the strongest outcomes were produced on the knowledge inference test which utilized progression visual fidelity accompanied by adaptive fading self-explanation prompts. In terms of what the results reveal about extraneous load, learners who had received adaptive fading self-explanation prompts reported having invested less effort and perceived the knowledge as being less difficult to learn as opposed to the learners who were provided with fixed self-explanation prompts. The results regarding germane load showed that learners who received progression visual fidelity condition concentrated more during learning, while those who received visual representations remained equally focused. Additionally, this study examined how cognitive load mediates between visual representation fidelity and mental models. The results showed that germane load partially mediated the beneficial effects of progressively realistic visual representations on knowledge inference.