AI’s wide open: the future of higher education

From personalised learning to behavioural nudges, artificial intelligence is set to transform HE for students says VitalSource’s Alice Duijser

Artificial intelligence (AI) has enormous potential in helping to address the higher education sector’s biggest challenges: falling lecture attendances, higher student expectations, and accessibility.

This comes at a time when the teaching excellence framework (TEF) is beginning to grade individual universities across the UK. TEF assesses teaching in universities and how well it ensures positive outcomes for students, both in terms of graduate-level employment and further study.

Digital leaders in higher education are rightly interested in the benefits AI tools can bring when it comes to maintaining high standards. Indeed, a recent survey by Jisc found that online learning tools and AI came first and second respectively at the top of their wish-lists.

Increasingly, we are seeing AI integrated into online learning tools. It’s a really exciting time for the digital publishing sector, and for the industry as a whole. Students can expect an online learning experience that is both personalised and adaptive.

‘Adaptive’ and ‘personalised’: these are two terms we readily associate with AI. But it’s important to understand what they actually mean in practice, the key distinctions between them, and the many benefits they can bring to lecturers and their students.

Lecturers will be equipped with robust learning data, providing them with much greater insight than previously possible

Personalised learning

One of the great advantages of AI is its ability to create a more personalised learning experience. Learning tailored to the needs and goals of each student can significantly help improve overall outcomes.

One example of personalised learning is ‘nudging’. The concept has gained an increasing amount of traction in business and politics. Initially developed by behavioural psychologists, it’s a simple way to positively influence behaviour. For instance, a supermarket might choose to place its fruit and vegetables at eye level, or near the checkout, instead of confectionary. In environments where this has been tested, a higher proportion of shoppers buy the healthier item. Consequently, the supermarket contributes towards healthy eating.

Similarly, nudging features embedded in digital textbooks can assist students in their learning by encouraging and reinforcing positive patterns of behaviour. Nudges can provide a student with subtle prompts directing them to areas of the course that they might be struggling with, or toward an area which has received insufficient attention. For instance, the AI system could recognise where a student is ignoring or missing helpful links guiding them to the right answer. At the end of a session, the system might nudge them towards clicking on a link, pointing out that students who do so tend to score much higher on assessments.

Personalisation also means providing students with targeted feedback, tailored to acknowledge the specific learning styles of individual students and issuing responses accordingly. This might mean feedback on completion and accuracy rates for assessments, or providing students with intelligently-deployed hints in areas students find challenging.

“One of the great advantages of AI is its ability to create a more personalised learning experience”

Adaptive learning

The adaptive aspect of AI tools can be defined as the way they modify how course materials are delivered through eTextbooks, based on students’ interaction with it. This includes the collective interactions of multiple users on the course.

In practice, this means that every time a student answers a question, it helps to inform the next piece of content they see. This means that students could be presented with different course materials and assessments, based on the way they best learn. Course material will adapt and change over time, refined through the everyday interactions of students.

This doesn’t mean to say that lecturers won’t have ultimate control over content. Core content can be made more or less adaptive, based on the lecturers’ preferences. They will be able to determine how much of a course should be considered core content, which all students have to learn, and what should be left to AI.

Additionally, lecturers will be equipped with robust learning data, providing them with much greater insight than previously possible. This includes better understanding of students’ study habits, their mastery of learning objectives, and those at risk of dropping off the course altogether.

Related feature: How to make AI in education ethical

Exciting times ahead for the sector

AI is high on the list of priorities for leaders in the higher education sector. They are interested in the way it can help address some of the sector’s challenges, improve overall outcomes for students, and raise overall standards of excellence.

At VitalSource, we’re proud to be a leading provider of AI-driven learning tools and in helping the sector take advantages of all the opportunities they bring.

Alice Duijser is MD of VitalSource