Using multimodal learning analytics to model students’ learning behavior in animated programming classroom

Abstract

Studies examining students' learningbehavior predominantly employed rich video data as their main source of information due to the limited knowledge of computer vision and deep learning algorithms. However, one of the challenges faced during such observation is the strenuous task of coding large amounts of video data through repeated viewings. In this research, we con rm the possibilities of classifying students' learning behavior using data obtained from multimodal distribution. We employed computer algorithms to classify students' learning behavior in animated programming classrooms and used information from this classi cation to predict learning outcomes. Speci cally, our study indicates the presence of three clusters of students in the domain of "stay active", "stay passive", and "to-passive". We also found a relationship between these pro les and learning outcomes. We discussed our ndings in accordance with the engagement and instructional quality models and believed that our statistical approach will support the ongoing re nement of the models in the context of behavioral pro ling and classroom interaction. We recommend that further studies should identify different epistemological frames in diverse classroom settings to provide su cient explanations of students' learning processes.

Exploring Attentional Dynamics in Animated Programming Environments: Trajectories, Variability, and Predictors

“…During the course of our investigation, we [5] first conducted a study that classified student attention-related behaviors using machine learning algorithms that disclosed important learning profiles. We were able to find three representative sequences that explain students' profiles when they were exposed to block-based programming activities as a supplement to their formal programming teaching.…”

Research on the patterns of the organic growth knowledge production based on a dual perspective of content and behaviour

“…Moreover, two coders discussed and encoded 200 units in the topic-coding stage and 235 units again in the behaviour-coding stage to reach a consensus. Use the LSA to explore the behavioural patterns around the same topic and the interactive mechanism of different generative topics LSA is applicable in analysing a series of observed events that involve sequential dependency and dichotomous behavioural cues that represent different system states (VanLear, 2017;Yusuf et al, 2023). The statistical significance of a specific behaviour followed by another can be examined by the LSA, and a diagram of behavioural patterns can be discovered and visualized (Bakeman & Gottman, 1997).…”