Adaptive learning technology allows children to learn at their own pace and helps prepare them for a future of living with artificial intelligence (AI). Kathryn Bates talks with Susanne de Mooij, whose research focuses on intelligent learning technologies, about the benefits and limitations of this technology in the classroom.

Kathryn Bates: What is adaptive learning technology, and how can it benefit children’s learning?

Susanne de Mooij: Adaptive learning technology is based on computerized algorithms that adapt learning materials to the ability of the student. In many classrooms, teachers have the challenging task of dividing their attention among as many as 30 or more children. Algorithms can assist by acting as a personalised tutor for each child. Adaptive learning technology recognises patterns in the child’s responses and then adapts to provide the personalised materials the child needs. The benefit is that each student is able to work at the appropriate level.

KB: I can see how this technology might benefit children who are improving, as they are able to move on to more difficult problems at their own rate. But what happens when a child is struggling or performing at the lowest level?

SdM: This technology is actually most beneficial in a class of children with different ability levels, where some might be struggling with a particular concept while others are thriving. The way I see it, a typical classroom with no adaptive learning technology is like a train that continues to move over the course of the year; children who are unable to hop on the train are left behind as it moves on without them. Even when a child fails to grasp a particular concept, a more advanced concept that relies on an understanding of the first one will still be taught to the class. This is where adaptive technology can help. It allows the child to practice at an easier level and introduces similar concepts that help the child meet the learning goal. It helps children get on the train.

“The way I see it, a typical classroom with no adaptive learning technology is like a train that continues to move over the course of the year; children who are unable to hop on the train are left behind as it moves on without them.”

In theory, this technology is good for every child. In practice, it is difficult to measure precisely how it helps struggling children. A lot of schools in the Netherlands, for example, use adaptive learning technology when children are struggling, and teachers report seeing benefits from the extra support. But we need more research to understand whether it is actually more beneficial for learners who are already performing well than it is for struggling learners.

KB: That’s interesting. With adaptive learning technology, struggling learners can spend more time learning a specific concept rather than trying to keep up with the class and moving on before they’ve learnt it properly?

SdM: Yes, exactly. There is a bit of a misconception that the system itself is a teaching method, but that’s not the case. Its purpose is always to support the teacher. Most of these systems include a dashboard that allows teachers to monitor the learning goals students are struggling to meet. But the system can’t work on its own. It’s only a computer; it can only do so much.

KB: How well does adaptive learning technology fit into classroom teaching? Do the students and teachers you work with find it useful?

SdM: In my experience, students enjoy using it and seem to learn a lot from it. Teachers tell us they like it, and they provide feedback on how well it’s working. But some teachers become frustrated by its limitations, and a handful of schools are against any kind of technology in the classroom.

“It can be difficult to strike a balance between what the technology can do and what teachers or students want.”

Some frustrations are difficult to avoid because of how the algorithms work. In the case of Math Garden, for example – an online learning environment for primary school children – the algorithm uses the speed of response in addition to accuracy to estimate how well the student is performing. Time pressure in this programme can be high; learners have just 8-20 seconds to respond to a question. Teachers want more agency over this aspect – they want the option of either including or removing the time limit, because the pressure can be too much for some students. That’s not possible, though, because the learning algorithm is dependent on time pressure. It can be difficult to strike a balance between what the technology can do and what teachers or students want.  

More on how AI can empower children’s learning
Giving children agency in an era of artificial intelligence

KB: Positive teacher-student relationships are important for both academic achievement and mental health outcomes. How can we ensure that increased use of this technology doesn’t compromise the relationship between teacher and student?

This is a really important question. Of course, we need to be careful to monitor the amount of time children spend in front of the screen and the activities they’re engaging in. The advantage of this technology is that it frees up attention and time that the teacher can then spend on more complex tasks, such as supporting student health and wellbeing. That being said, we need to make sure that teachers know how to limit time spent on this technology, because the companies that produce it are unlikely to impose limits.

KB: I guess there will always be a need for compromise, from the teacher or the technology. So what is the best use for this technology in the classroom?

SdM: AI technology is currently providing an accessible environment to which important tasks, such as monitoring progress or evaluating whether to proceed to new learning goal, are offloaded from the teacher. We’re still in the early days; these systems have only been around for about 20 years. As the technology continues to be developed, we hope that students will have greater agency over their learning within the technology. They should not become dependent on AI; instead, they should learn to work with it as a tool that allows them to progress and thrive.

“As the technology continues to be developed, we hope that students will have greater agency over their learning within the technology.”

Footnotes

Susanne de Mooij is a postdoctoral researcher in the Learning & Plasticity group of the Behavioural Science Institute and a member of the Adaptive Learning Lab (ALL) at Radboud University in Nijmegen, Netherlands. Her research seeks to optimize intelligent learning technologies to enhance the learning experiences of individual students. She is interested in how science and the edtech industry can collaborate to have a positive impact on students’ learning. De Mooij is a postdoc at CELLA (the Center for Learning and Living with AI), funded by the Jacobs Foundation.

Twitter: @SusanneMooij

Keep up to date with the BOLD newsletter