How Are Higher Education Institutions Using Analytics?

How are universities using analytics to help improve their results?


August 13th was A level results day in the UK, and for undergraduates, they are looking at which higher education institution to attend. Universities are now looking to Big Data to leverage a variety of insights that can benefit both themselves and their students - improving student acquisition, lower drop out rates, and helping to raise performance.

Retention rates are vital to maintaining a university’s reputation. According to the Higher Education Funding Council for England (Hefce), more than 8% of undergraduates drop out in their first year of study. This costs universities roughly £33,000 per student. Analytics can look at a range of metrics to reveal signs that a student may be about to drop out, primarily by looking at student engagement and comparing this with past data from previous drop outs. Universities are looking at how often a student goes into the library, how many books they take out, and how often they sign into their virtual learning environment. Teachers and administrators can then tailor the response to the student by looking at any deviations and picking up on similarities, using personal data to discover things about them so they can adjust their protocols to ensure the student does not drop out. For example, London South Bank University discovered that once something went wrong for non-traditional or international students, noting that ’the downward spiral was far quicker than for other students, so the earlier in that cycle we could catch them the better the results.’ By utilizing predictive analytics, they are now able to respond at the top of the spiral.

Student engagement is also connected to improving performance. Nottingham Trent university found that students whose engagement was 80% or more got at least a 2:1. They’re also examining previous tests results as well as clustering students so that teaching methods can be adapted to suit different needs. By looking at personality metrics, professors can tailor class sizes and increase the number of seminars or lectures as appropriate. Nottingham Trent University’s student engagement manager, Ed Foster, noted for example that: ‘What we saw, is that if you’re from a BME or low socio-economic background, then participating is a far more important factor in your progression than your background.’

Nottingham Trent have even adapted their teaching methods to use something called ‘flip learning’. Flip learning is a way of teaching, whereby students do their homework before a class, and the class is then used to focus on any issues and challenges they faced in doing it. They discovered that this is actually more beneficial to students than traditional methods, in which a class is taught and then the homework is set afterwards.

There are issues that must be kept a check on. Students are notoriously unengaged, and liable to try and game the system if they know that they’re being checked up on, for example by taking library books out and then not reading them, simply to appear engaged. In order to prevent this, it is necessary to use a variety of metrics in conjunction with one another. 


Read next:

Working At The Boundaries Of Aesthetics And Inference