2020/07/14 by Robert L. Peach, Peach, Robert L., Sam F. Greenbury +9
Computer Science · Social Sciences · #Computers and Society (cs.CY) #FOS: Computer and information sciences #Intelligent Tutoring Systems and Adaptive Learning #Online Learning and Analytics #Online and Blended Learning #Social and Information Networks (cs.SI)
paper · pdf · doi:10.48550/arxiv.2007.07003
openalex publication_date 2020/07/14 · openalex created_date 2025/10/10 · openalex updated_date 2026/07/28
The intrinsic temporality of learning demands the adoption of methodologies\ncapable of exploiting time-series information. In this study we leverage the\nsequence data framework and show how data-driven analysis of temporal sequences\nof task completion in online courses can be used to characterise personal and\ngroup learners' behaviors, and to identify critical tasks and course sessions\nin a given course design. We also introduce a recently developed probabilistic\nBayesian model to learn sequence trajectories of students and predict student\nperformance. The application of our data-driven sequence-based analyses to data\nfrom learners undertaking an on-line Business Management course reveals\ndistinct behaviors within the cohort of learners, identifying learners or\ngroups of learners that deviate from the nominal order expected in the course.\nUsing course grades a posteriori, we explore differences in behavior between\nhigh and low performing learners. We find that high performing learners follow\nthe progression between weekly sessions more regularly than low performing\nlearners, yet within each weekly session high performing learners are less tied\nto the nominal task order. We then model the sequences of high and low\nperformance students using the probablistic Bayesian model and show that we can\nlearn engagement behaviors associated with performance. We also show that the\ndata sequence framework can be used for task centric analysis; we identify\ncritical junctures and differences among types of tasks within the course\ndesign. We find that non-rote learning tasks, such as interactive tasks or\ndiscussion posts, are correlated with higher performance. We discuss the\napplication of such analytical techniques as an aid to course design,\nintervention, and student supervision.\n