Universities and colleges wishing to maximise student success and raise achievement of all students – including those from non-traditional backgrounds – should consider adopting learning analytics, as recommended by the Higher Education Commission in its recent report ‘From Bricks to Clicks’.
Whilst this is a relatively new area, a growing number of potential learning analytics solutions are available. It’s becoming an increasingly competitive and innovative marketplace, as captured in a recent Jisc review into the current state of play in the UK.
Choosing a system
Investing in the right digital solution can be challenging for universities and colleges at the best of times, but more so in the early stages of a new digital solution marketplace, where price points are less-well established and the risk of vendor lock-in is higher.
Key to making the right choice is being clear on the particular educational goals your solution is intended to address. For example, some organisations may wish to focus on completion rates and student retention, others on improving academic achievement in a particular domain (such as firsts and 2:1 borderlines), and others on yielding data to shape and enhance the content of the digitally-enriched curriculum.
While such individual, tailored approaches are entirely appropriate for a sector that values diversity and institutional autonomy, as well as a growing educational marketplace, there is a risk that the broader benefits of a joined-up approach may not be realised.
Such joined-up benefits for universities and colleges might include: economies of scale; better understanding of solution pricing; mitigation against lock-in; ability to benchmark against peers; general ‘de-risking’; and for vendors: reduced cost of sales; and lower barriers to innovation.
Is there some way we can get the best of both worlds in the UK, balancing both individual and joined-up approaches? And that benefits universities, colleges and vendors – a genuine win-win? At Jisc, we think there is…
Solution architecture and framework for the sector
Ever since it was highlighted in our co-design process as a priority area by our members, we at Jisc have been seeking to ‘make the market’ for UK learning analytics, by working with universities, colleges, and leading vendors, to define and implement a national open architecture for learning analytics.
At the centre of this architecture is a secure, shared, cloud-based multi-tenanted learning analytics data warehouse, which aggregates two main data types from local learning-related information systems:
- A ‘small data set’ of canonical student record data via a new universal data model, eg student’s course, modules, grades, etc.
- A ‘big data set’ of detailed student digital activity, eg from virtual learning environments (VLEs) like Moodle or Blackboard, interactions with library systems, and self-declared data via a student app, all translated and stored in a format called xAPI
Where we are now
Right now we have a beta-stage national learning analytics service, which we’ll be piloting with 50 organisations over the coming year and that we believe is globally unique.
By mining aggregated data in the national warehouse (and with the strict permission of data owners in all cases), we aim to create new canonical metrics that, for the first time, will allow meaningful comparison of analytic patterns in student engagement and other proxy measures for teaching quality.
Alongside, via openly-defined APIs, vendors are able to securely interface their analytic dashboard and predictive tools to sections of the national warehouse, seamlessly providing the tailored services that individual universities and colleges require in pursuance of their own distinct missions and organisational goals.
Working with the vendor community
Now, we want to open up our architecture more widely, to any vendor who wishes to collaborate and engage. We recently hosted a vendor event, where over 30 analytics product vendors met the Jisc team, and learnt how they could get involved.
Civitas Learning is one of a number of additional vendors now working closely with us as a result of the event. A leading analytics organisation in the US, Civitas has experience in using advanced analytics, machine learning and data science, through a vendor and solution-agnostic platform that’s integrated with more than 93% of VLEs and 85% of student information systems. It produces multiple institution-specific models, building approximately 30 predictive models per institution to more precisely leverage learning analytics. In the UK Civitas Learning’s work now spans post-1992 universities to Russell Group institutions, and its network currently includes more than 285 institutions campuses reaching more than 6.5 million students across the world.
If you’re a vendor who like to get involved please contact Michael Webb, our director of technology and analytics.
Colleges and universities hoping to find out more about learning analytics should speak to their account managers.