Predictive modelling for addressing students’ attrition in higher education: The case of OU Analyse

Cloud created by:

Rebecca Ferguson
15 May 2017

Christothea Herodotou, Alison Gilmour, Avinash Boroowa, Bart Rienties, Zdenek Zdrahal and Martin Hlosta

Learner retention is a critical issue in distance education. A number of strategic initiatives are taking place at the Open University (OU) UK, both internal and external, focusing activity on the development of approaches to enhance student outcomes. This presentation will detail one institutional intervention, the OU Analyse (OUA), which draws on predictive modelling and student probabilities, to identify and predict students at risk of failing their studies. OUA (https://analyse.kmi.open.ac.uk/) uses machine learning methods to improve student retention. It provides weekly early warning indicators of students who may be at risk of not submitting their next assignment. It is intended to be used by tutors and student support services to enhance student retention across the OU. It was originally piloted with 10 courses. A follow-up evaluation has been planned with 25 courses across all faculties to conclude on its impact on student performance. Findings from the piloting of OUA with 10 modules will be presented. Data were collected from students' completion, pass and withdrawal rates and semi-structured interviews with tutors who used OUA. This presentation aims to provide illuminating insights about the use of predictive modelling in Higher Education (HE) and spark discussions about how institutions can design and implement effective interventions for minimising early exit from HE and enhancing the learning experience.

 

Extra content

Embedded Content