There’s been a revolution in healthcare. The patient at the heart of the system is now an accepted part of healthcare design. That’s not for a sense of the warm-and-fuzzies but a response to the challenges of (severely) constrained resource and a need to do more, with less. Understanding patient behaviour is a key part of the transformation: how can systems work with – rather than against – patients’ instincts in order to be more efficient? One key element of that is the psychology of health behaviours; another is the use of data to analyse problems and design creative solutions.
Why can’t we apply the same approach in universities? Most institutions are gathering – or have the capacity to gather – data all the time on student attendance, academic performance and the location of the student is via their wifi logins, cashless spending cards or building access. And most providers are capable of analysing patterns of thousands of past and current students. With all that data at hand, it’s possible to look for the trends which can predict likely outcomes. Just as we want to prevent a patient from taking medicine incorrectly, surely we want to identify students who are at risk, say, of dropping out in order to intervene.
“There’s a massive opportunity to draw together data from across the silos in universities,” KPMG’s Simon Livings says. “If you can combine the data and identify the warning signs of student disengagement, then you can put the right intervention in place.” That idea of a ‘right intervention’ wouldn’t simply mean some kind of automated call or WhatsApp message, but also the services which universities already provide but which may currently rely on a student’s self-referral. Automating systems, and using data to support student success, should bring savings as well as improving outcomes.
Why isn’t it happening?
There are also obvious question marks around the important and knotty questions of consent and ethics. Is it right for a university to use information about a student’s movements gained from tracking their phone across campus? Do we have the right safeguards in place to ensure that consent is always properly informed and that data won’t be misused? How far should we partner with students themselves – as the NHS has done very successfully with patients – to engage with the right interventions?
On Wonkhe last year, a strong case was made that: “those of a more conspiratorial outlook, with the best of intentions no doubt, may inadvertently be delaying progress rather than supporting the safe and considered development of what may prove a powerful approach to fostering student success.” I wouldn’t argue for us to ignore the questions of ethics, good governance and data security. But those questions shouldn’t prevent us from having a considered and productive debate about the use of analytics to improve the services provided to students.
There are internal obstacles to overcome, as Paul Henderson, a healthcare expert at KPMG, says: “The barriers to deploying more creative data analytics are almost always down to demonstrating return on investment. Finding the money to invest in technological solutions can be difficult. But even in the cash strapped NHS, the savings are there if you can use predictive analytics ethically, securely and effectively.” As well as improving the student journey, more creative data analytics can be used to improve reporting, forecasting and scenario planning. That should, if done well, reduce risk and improve efficiency.
While health might offer a good comparison with education, it also highlights just how far behind most of higher education lags when it comes to using data analytics to improve the ‘customer’ experience. And that’s where the focus needs to be: this isn’t about selling more widgets, but about investing in technology to improve education outcomes. That could be to support students who might otherwise struggle in HE or about retention – which doesn’t just help a university’s bottom line but also the student.
How many universities have a complete overview of their multiple systems, how they interact, and a good sense of what can be done with their data? From outreach to recruitment to admissions, and throughout the course (and beyond), does the institution know exactly where the right intervention will have most positive impact? Those might like questions rather too big for complex organisations, but big businesses and public services are hardly less complicated.
The use of data is only likely to grow in importance. The volume and speed of data collection will surely increase and with that the opportunities it represents. Competition between providers will further differentiate those who have the command of the data, and those that have fallen behind. If a case can be made which supports more efficient and cost-effective – business delivery, as well as improving student outcomes, that seems like a no-brainer to me.