No Students Left Behind: Unlocking the Potential of Predictive Analytics in Advancing Fair and Accurate Educational Decision-Making

This thesis explores how predictive analytics can support student learning while addressing the risk of bias against certain groups. It goes beyond focusing on overall accuracy to ensure fairness is included in these tools. The research applies predictive models across different educational settings and aims to develop methods that are both accurate and fair. By doing so, this work seeks to build trust in educational technologies and promote their use throughout various stages of students’ academic journeys.