- This HEPI blog was kindly authored by Simon Rimmington, Director of the Foundation Year at Keele University. His piece is a case study of how a data-driven approach can support student success.
In the rapidly evolving landscape of higher education, universities are increasingly embracing engagement dashboards and learner analytics as powerful tools in their quest to enhance student success. Internal and external drivers are propelling this shift, with academic institutions adopting early warning systems to efficiently identify students at risk.
At Keele, the Foundation Year has seen significant growth in student numbers over recent years, with approximately 750 students joining for 2023/24. In order to foster a supportive environment between students and their academic mentors, we are using real-time engagement data provided by StREAM to implement more nuanced and holistic approaches for individual students.
Moreover, each student is assigned an Academic Mentor (AM) and access to a Student Experience and Support Officer (SESO) who works across the FY and Student Services.
Integrating the Dashboard Pilot
A successful pilot of the StREAM dashboard, of which the foundation year was an integral part, allowed us to build confidence in the engagement profiles of this cohort of students.
Building on this leap forward in data visibility, in 2022/23, the dashboard was implemented across the university, with all academic mentors receiving training on how to use it. The data, and other indicators, such as assessment submissions, were used to identify and support students who might need academic assistance. Subsequently, during 2023/24, we have improved our understanding of the engagement profiles and introduced several interventions targeted at student cohorts based on their profiles.
Although the data is compelling, the next challenge is to turn this knowledge into positive action. During the pilot, we saw that 73% of students with a ‘High’ or ‘Good’ profile for the semester passed all their modules, with a further 10% needing reassessment in one or two assessments. Conversely, only 10% of students with a ‘Low’, ‘Very Low’ or ‘None’ engagement profile passed all their modules with 53% failing all modules. These students tend to be hard to reach and there are potentially complex issues behind their low engagement. Therefore, interventions that are timely and relevant in providing the appropriate academic or personal support are key to a positive impact.
At an individual level, the student engagement dashboard aims to provide a shared platform for students and their academic mentors to engage in more meaningful and holistic conversations. By granting students access to their engagement data, the dashboard can help foster a sense of transparency and empower students to take ownership of their academic journey, encouraging a shift towards more comprehensive and insightful discussions.
The ability to see a student engagement profile in real time enables mentors to track a student’s engagement patterns over time. The intervention record associated with specific engagement metrics could help to facilitate in-depth discussions and aid more personalised feedback. The inclusion of a visual representation of engagement data could ultimately empower mentors to guide students in identifying trends and patterns and uncover underlying factors that may be influencing a student’s progress. This proactive approach helps to ensure that students receive timely support and guidance to navigate challenges and achieve their academic goals.
Cohort-level support/interventions
To support the academic mentoring system, we introduced a beta-phase cohort-level approach to identifying students at risk. The Keele Reporting Dashboard uses the data from StREAM to provide cohort-level data that can be filtered to build up a database of engagement for all the FY students over time, identifying engagement trends or changes for individual students.
Over the semester, a range of communications and interventions were used. These communications were supportive but queried reasons for ‘Very low’ engagement using emails, Teams messages and MS Forms. Attendance data was also included and has been used to nuance communications. For example, “We can see that your engagement is Low, but your attendance is quite good. Are you having problems accessing materials?”.
Revealing patterns in disengagement
As the semester progressed, several persistent, very low engagers received communication from their Senior Academic Mentor around the risks of withdrawal and the subsequent consequences of withdrawal; importantly, these interventions included support to plan the next steps, e.g. a leave of absence or a study, plan for the rest of the academic year. By creating this process founded on data, we have seen a reduction in non-continuation from 6.8% by the end of semester 1 in 2022/23 to 3.4% in 2023/24 in the same period.
Regardless of the underlying causes for disengagement, prolonged inactivity risks perpetuating a downward spiral, fostering negative sentiments toward university life. It is essential that some form of ‘reach-out’ to students is initiated as early as possible, and this should be characterised by being supportive, open, and honest. This is underpinned by the StREAM learner analytics platform, and although we are only in the second year of deployment, having better tools, developing and evaluating our approaches provide a greater sense of empowerment and success. In general, we can identify positive movement in improving student retention and ultimately, success. To see more of the findings relating to the Keele Foundation Year, please read the full case study.