Artificial Intelligence (AI), there is a lot to be done. Proponents see in AI the solution to many of the problems facing our education system. With AI, we can offer students truly personalised education. AI can reduce the workload of lecturers. The possibilities seem endless. Opponents mainly see risks. AI could send students in the wrong direction. AI takes over important core tasks of the lecturer. I don’t think the use of AI can be that positive or negative. There must be possibilities to use this powerful technology in a useful and effective way. Work evidence-informed, is my suggestion.
I am always curious when I read about a successful application of AI in education. From my background as a medical biologist and lecturer in the science domain, examples from this domain quickly catch my eye. For example, I was recently triggered by the presentation of the higher education prize 2021 to Amsterdam University of Applied Sciences for the appLeerlevels. This app uses AI to help secondary school students learn physics in a personalised way. It gives students the time and space to learn at their own pace and the lecturer does not have to keep explaining the material in class, but can match the level of each student. Wonderful! And why does this work so well?
The basis of the Leerlevels app is a very finely meshed, ordered network of learning objectives. In-depth knowledge of subject matter is therefore required. The app also provides constructive feedback. How do you ensure that this feedback is really effective and encourages learning? This requires knowledge of cognitive behavioural sciences. I’m trying to imagine the steps that have been taken and how much work it has been to arrive at this app. All the (literature) research into subject teaching methods, the arrangement of learning objectives and the constant breaking down into smaller learning objectives, the testing of formulations. And then finally the first small test. All those small steps that eventually led to this app. A fine piece of work.
Leerlevelsshows very well how AI can be used to support education in the natural sciences. The logical connection between the learning goals makes it possible to determine very well where a learner stands, and which learning goals need extra attention, which the programme then intelligently responds to. But how can we use AI in fields where the relationship between the learning objectives is less clear? Or in the guidance of students? Here, too, AI offers possibilities. More and more, self-learning AI is being used to find logical connections through data analysis and to build a network that can then be tested. This requires large amounts of data and that is where the challenge lies. How reliable is the data? How representative? How robust is the resulting model? Does it still fit, for example, if the composition of the student population changes?
AI has enormous potential and it is this potential that makes trust in the use of this technology vulnerable. Do we want to develop and use AI applications with the trust of students, lecturers and society? Then we must be able to explain again and again why the use of AI is appropriate in a particular context, be transparent about how it works, guarantee that study data is used safely and reliably, and constantly evaluate whether the application contributes to the intended goal. By making evidence-informed working – the use of data and research results as a starting point in decision-making – the standard in the development and deployment of AI in education, we are actively implementing this.
The zone evidence-informed educational innovation with IT provides tools, examples and instruments that help to innovate and implement IT in education in an evidence-informed way.
June of the Month of AI in Education. A month full of webinars, hackathons and blogs. Knowing more? Look in our calendar.
Photo by Markus Spiske via Unsplash
Share this page
Want to get started with innovation within your institution? The Evidence-informed educational innovation with ICT zone and the Digital education
What factors influence the evidence-informed educational innovation process within higher education institutions? With this as its central research question, the
If you use instruments for evidence-informed educational innovation with IT within your higher education institution, you should describe them using
Good practices of evidence-informed educational innovation with IT are good examples of evidence-informed ways of working within innovation projects or
