Designing Human-Centred AI-powered Learning Analytics

This thesis explores how Artificial Intelligence and Learning Analytics can be designed for education in ways that are safe, reliable, and trustworthy. It adopts a human-centred approach, engaging teachers and students in shaping systems that genuinely support learning and reflection. By drawing on different types of data and collaborating throughout the design process, the research develops and evaluates methods that assist teachers and students to reflect, improve practice, and adapt systems through real-world use. The findings demonstrate that, when the design of AI tools is guided by human expertise, they can foster trust, encourage adoption, and support teaching and learning in meaningful, practical ways.