Developing an open-source learning analytics tool for providing insights to support students and improve teaching practice

Learning analytics and data-driven approaches have gained real traction in higher education institutions, enabling insights into student attendance, engagement, and performance trends. This paper introduces the development of an open-source course-level data analytics tool, 'the student record,' designed to complement the conventional institutional approach with a responsive, context-specific, and user-centered method for gathering, analysing, and presenting student attendance data. In developing the tool, our team experimented with Microsoft Excel's natural language interface and artificial intelligence capabilities and currently the tool uses predefined python scripts in a Jupyter Notebook to generate visual dashboards, flags, and recommendations based on statistical process control (SPC) principles and machine learning models. This approach aims to facilitate a dynamic student engagement dialogue between staff and students, fostering a rolling academic health check to support students and provide contextual insights for module teams and course leaders. The tool's ethical considerations prioritise transparency, data privacy, and student accessibility to their own data, aligning with personal academic tutoring systems. The paper outlines the tool's development, capabilities, and a critical discussion of key aspects, including AI integration, data security, and ethical implications.