Tracing learning strategies in online learning environments: a learning analytics approach

Learning has expanded beyond formal education; yet, students continue to face the challenge of how to effectively direct their learning. Among the processes of learning, the selection and application of learning tactics and strategies are fundamental steps. Learning tactics and strategies have long been considered as key predictors of learning performance. Theoretical models of self-regulated learning (SRL) assert that the choice and use of learning tactics and strategies are influenced by the internal (cognitive) and external (task) conditions. These conditions are consistently updated when students receive internal/external feedback. However, internal feedback generated based on students’ evaluation of their own performance against the expectation and/or learning goal is not accurate. Guiding students to apply appropriate learning strategies i.e. providing external feedback, hence, could enhance the students’ learning. Recent research literature suggests that learning analytics can be leveraged to support students in the selection and use of effective learning tactics and strategies. However, there has been limited literature on the ways this can be achieved. This thesis aims to fill this gap in the literature. This thesis begins by exploring the state of the art regarding how students receive learning analytics-based support for the selection and application of learning tactics and strategies. The systematic literature review on this topic reveals that students rarely receive feedback on learning tactics and strategies with learning analytics dashboards. One of the barriers to providing feedback on learning tactics and strategies is the difficulty in detecting learning tactics and strategies that students used when interacting with learning activities. Hence, this thesis proposes a novel analyticsbased approach to detect learning tactics and strategies based on digital trace data recorded in learning environments. The proposed analytics-based approach is based on process, sequence mining and clustering techniques. To validate the results of the proposed approach and the credibility of the automatically detected learning tactics and strategies, associations with academic performance and different feedback conditions are explored. To further validate the approach, the efficacy of each proposed approach in the detection of learning tactics and strategies is investigated. In addition, the thesis explores the alignment of the automatically detected learning tactics and strategies with relevant models of SRL. This is done by examining the association between the internal conditions and external conditions. Specifically, internal conditions are represented by the disposition of students based on self-reports of personality traits, whereas external conditions are represented by course instructional designs and delivery modalities. The thesis is concluded with a discussion of the implications of the proposed analytics methodology on research and practice of learning and teaching