No abstract

Despite the potential of Predictive Learning Analytics (PLAs) to identify students at risk of failing their studies, research demonstrating effective application of PLAs to higher education is relatively limited. The aims of this study are 1) to identify whether and how PLAs can inform the design of motivational interventions and 2) to capture the impact of those interventions on student retention at the Open University UK. A predictive model -the Student Probabilities Model (SPM) -was used to predict the likelihood of a student remaining in a course at the next milestone and eventually completing it. Undergraduate students (N=630) with a low probability of completing their studies were randomly allocated into the control (n=312) and intervention groups (n=318), and contacted by the university Student Support Teams (SSTs) using a set of motivational interventions such as text, phone, and email. The results of the randomized control trial showed statistically significant better student retention outcomes for the intervention group, with the proposed intervention deemed effective in facilitating course completion. The intervention also improved the administration of student support at scale and low cost. Notes for Research• This paper explains how predictive learning analytics can help with the selection of students who need support.• A motivational intervention delivered via text, phone, and email was tested.• Students receiving the intervention had statistically significant better retention outcomes compared to the control group.• Student Support Teams could use predictive learning analytics to target specific students and better support them.