Data-driven decisions: a conceptual framework for integrating educational data mining in professional bodies

This thesis investigates the transformative potential of Educational Data Mining (EDM) in shaping educational landscapes and enhancing student outcomes. Focused on a comprehensive exploration, the research delves into multiple dimensions of EDM, addressing critical areas such as student dropout prediction, identification of high-performing students, and the nuanced characteristics associated with academic excellence. The overarching goal is to provide actionable insights for institutions and educators, facilitating informed decision-making in areas like student recruitment, course success prediction, and fostering sustainability and equity in higher education. In doing so, the study contributes insights into how institutions can strategically refine their recruitment processes, identify and nurture key attributes of high-performing students, adopt robust and interpretable EDM methodologies for predicting academic success, and leverage EDM to ensure their approaches align with principles of inclusivity, equity, and sustainability.

The study formulates specific research questions, exploring how institutions can make strategic decisions during recruitment, the key attributes defining high-performing students, methodologies employed in EDM for student identification, and the predictive power of emphasizing sustainability and equity in educational practices. The findings contribute to the development of a robust conceptual framework that serves as a practical guide for institutions offering professional qualifications outside the traditional university model. This framework is designed to empower institutions to harness the full potential of EDM, ensuring its effective integration into diverse aspects of educational practices. The research employs a time-based sampling method, concentrating on a case study of an organization, spanning the years 2017 to 2022. The global scope of the study involves a dataset of approximately 150,000 registrants. The inclusion of diverse criteria such as gender, age, location, and exemption status enhance the representativeness of the dataset. However, it is acknowledged that certain limitations may arise due to potential data gaps, especially in older repositories.

The literature review undertaken in this thesis encompasses studies related to the identification of high achieving students, the exploration of characteristics contributing to student success, evaluation of predictive models, and an examination of the role of EDM in fostering sustainability and equity in education. The conceptual framework visually represents the workflow of the study, addressing key components such as streamlining student intake, identifying high-performing students, methodologies for achiever identification, and the pivotal role of EDM in promoting sustainability.

The thesis concludes by summarising key findings drawn from an extensive analysis of anonymised student data, emphasizing the importance of factors such as first exam performance, persistence, and preparedness in student success. It underscores the evolving dynamics of student progression revealed through EDM insights, providing a nuanced understanding that can guide educational institutions towards more effective practices and outcomes.