A theme that emerges from a number of the interviews with experts in learning analytics is that adaptive learning systems and intelligent tutoring will continue to improve and become increasingly important in education. Alan Berg believes that, after tackling retention, these systems should be the next focus for institutions, enabling misconceptions to be spotted much earlier, with machines feeling increasingly like humans. Josh Baron agrees that adaptive learning systems may be the next major development:

I think we’re scratching the surface right now with pretty much all the learning analytics stuff. Academic early alert systems may be where most people are focussed but I think even there … it’s still early days. I would say that adaptive learning systems and cognitive tutoring technology, which I categorise as learning analytics at least, is the one that I think holds the biggest potential for directly impacting on learning, but I think hasn’t really left the lab yet too much.

And I think everything from improving learning outcomes to reducing the cost of delivering learning and creating more personalised learning – I think those types of adaptive learning systems hold the potential to do that kind of thing, and that’s why I’m really excited about that.

Abelardo Pardo’s vision, too, is that we should move beyond visualisation and go

all the way to true adaptation, true modification of the learning environment … providing students with recommendations or adaptions that go beyond simple, basic feedback, and focus more on higher level types of constructions, like, for example, learning strategies.

He thinks the systems will increasingly be able to provide students with meaningful suggestions, rather than simply offering new resources or exercises. Mark Milliron, meanwhile, talks about the ‘power of the learning moment’ and describes how this can lead to greater connections between individuals rather than simply replacing teachers with machines:

There’s a moment where you can say, ‘here are the right learning resources, here are the right interactivities, here’s a set of peers who just wrestled with this, you might want to connect with, here are a set of mentors who are already in the field who have connections on this’. We can bring to that learning moment the right resource at the right time in the right way … And sometimes that … means you’re going to have to turn the agent off and spend the next hour with this text, dive deep into it and then let’s talk about it and let’s engage. It does not mean spoonfeeding. The data can point to the fact that you actually need to talk to another person, you need to connect with a peer or a faculty member. That’s one of the most exciting parts of this, we can build on the fly communities … which can be powerful for students.