Designing for Self-Regulated Learning: A Dual-View Intelligent Visualization Dashboard to Support Instructors and Students Using Multimodal Trace Data in Classrooms

Effective learning analytics dashboards (LADs) should offer both instructors and students valuable insights into student learning. Information about one’s learning, as both a product and a process, should be provided to the user as actionable data visualizations and aggregated indicators of learning. However, most dashboards often solely focus on the instructor’s perspective, neglecting the impact of providing students with their data in a student view. Furthermore, these dashboards assume instructors are proficient in data analytics and can quickly interpret complicated visualizations in-situ while accounting for context and conditional factors. This challenge is further exacerbated by the lack of theoretically informed learning analytics principles and design choices as many of the dashboards rely primarily on performance-based data, neglecting process and trace data of cognitive, metacognitive, affective, motivational, and social (CAMMS) processes. As such, we are introducing MetaDash, a multimodal self-regulated learning (SRL) dashboard with both an instructor and student view. This dual-view (i.e., instructor and student-facing) dashboard prototype is populated with aggregate and contextualized multimodal trace data grounded within models of SRL, affect dynamics, information processing, cognitive load, and multimodal learning analytics. In this paper, we leverage ideas derived from SRL to identify gaps in current learning analytics dashboards. We then present the design principles and architecture of MetaDash, including how we derived its structure and how it supports our framework. We discuss the affective dynamics and learning analytics on the landing page to understand user engagement and detail how analytics are customized for different phases within the architecture. We highlight the advantages of incorporating real-time data analysis for immediate decision-making. Future research will focus on refining MetaDash through enhancements informed by user and focus group testing, experimental studies, and the integration of user feedback to address challenges and expand the dashboard’s functionality and effectiveness.