Our roundtable at THE Live discussed how higher education providers can utilise data to maximise students’ attainment, retention and well-being. The roundtable was co-chaired by Simon Baker, data editor at Times Higher Education, and our own Karen Rawson, senior product manager for further and higher education.
As universities look ahead to the data-led future, much discussion has taken place about the challenges involved in making valuable and ethical use of the information garnered on student enrolment, assessment and experiences.
Opening the discussion, Ms Rawson stated that, while data reporting is at the forefront of university leadership agendas, “it’s becoming very apparent that there’s a divergence between data quality, data validation, data entry, data statistics...between different segments of staff”.
“Is it possible to break down data silos?” she asked the panellists, “and is there a data intolerance with some academics?”
With data protection and privacy laws increasingly strict – and for good reason – participants agreed that fear of data misuse was a likely barrier for some departments and individuals when it came to using data resources to their institutions’ advantage. Academics and staff administrators were becoming increasingly wary about use and access to student information in particular, it was noted.
But a key ambition that quickly emerged was a shared desire to harness internal databases for the purposes of student safeguarding and well-being.
“The single most useful aspect of collating data is when it comes to students at risk,” said Katie Drapes, head of standards and quality enhancement for Regent College London. “Student retention is the one thing that everyone cares about – it seems to be one area that everyone can focus on with the same level of intensity, with the same level of outcome.”
As the quality of data management systems available to universities improves, it is hoped that campus support services will be able to pinpoint individual students in need of extra support or who are at risk of dropping out of their course.
But the panellists raised concerns about how to go about this ethically – with some stressing that targeting students in such a way could have a negative effect. A computer algorithm might flag an individual from a socially disadvantaged background as predisposed to struggle academically, for example, when that might not be the reality of their experience at all.
“Students don’t always want a reminder that universities don’t have faith in their abilities,” said Claire Chalmers, associate director of student recruitment at the University of Greenwich. A more “subtle approach to use of data” was required, she stressed, “approaching students, asking them how they are…but even then, they can still see through that”.
Several panellists commented that it was not data silos that presented the biggest barrier to their work but the limited quality of the data collected, alongside a lack of understanding among many university staff and academics of what to do with it.
A common misconception, explained Amy Smith, associate director of admissions and applicant engagement at Nottingham Trent University, was for staff to think “we’ve got access to all this data [so] we can make it say what we want. But, quite often, the quality varies,” she said, “Higher Education Statistics Agency data doesn’t talk to the Universities and Colleges Admissions Service data, and so on.”
“The biggest problem is not silos, but that it depends on how well resourced you are to use your data,” added Kieron Broadhead, executive director of student experience at the University of Southampton. Universities were in a situation where they had access to “an awful lot of data” after decades of collecting their own, as well as having access to new public figures from the Office for Students.
“There is an expectation that you will do lots with it,” he continued, “but there isn’t a sensible way of using that data asset and that’s because we don’t always know the question we want to answer. It’s a massive university asset that we don’t quite understand.”
Another major sticking point for the sector is a lack of clear communication between institutions and the external governing bodies that universities are required to report to, panellists suggested.
Following the creation of the Office for Students, universities in England are now required to submit “access to participation” data to the governing body, detailing how they aim to improve opportunities for under-represented groups in their own university community. It forms part of a plan outlined by the UK government to eliminate the gap in university entry rates between the most- and least-represented groups by 2038-39.
But panellists raised the issue of data submissions being returned and rejected by the body as a result of data inconsistencies and “unclear questions being asked of universities” – which meant that some institutions were having to submit answers three or four times.
“Universities are expected to act like analysts when that’s not their role,” said Ms Drapes. “We have to manipulate the way we report the data to fit the very fixed statutory requirement.”
As a consequence, such mandatory data-sharing exercises were creating silos in themselves, the group argued, which not only failed to improve communications but added to institutional burden.
Asked for ideas on how to improve the use of data when it came to student health and well-being, one suggestion to come from the roundtable discussion was for students to be given a national identification number, similar to the national insurance numbers given to citizens in the UK.
This would enable vital safeguarding information to be kept on record for students throughout their educational years in the same way that NHS employees carry their data from job to job.
“Data collection brings so many opportunities,” said Mr Broadhead. “Tailoring student support services, eradicating the higher education participation gap: it’s all possible if we go about it the right way.”