Digitalised students?
Renate Schubert thinks about how to approach students’ digitalised data – and sees a lot of potential, but also some open questions.
University teaching has started to produce huge mountains of data. More and more courses use electronic platforms like Moodle and OLAT, which can record which students log in to work through the course material and how often they do so; how long students spend on particular exercises; how many attempts they need to solve certain tasks correctly; and more. This data can produce plenty of interesting information, and it can also help students as well as lecturers to design a more efficient and successful learning process. So far, so good... but a few questions arise when you look more closely.
What happens when data is manipulated?
As a lecturer, I can use the data to get an impression of whether my exercises and tasks are too hard or too easy, and can react accordingly. I can plan to repeat certain information, or prepare additional or different material. But as lecturers, how can we be certain that the students are not using the whole thing strategically and producing data that leads us to believe, for example, that there are problems with the teaching material that don’t actually exist?
Are we limiting students’ motivation?
Our students can also benefit from this data. For example, they can compare themselves to their fellow students and evaluate their own learning success and behaviour. Ideally, the students would then put in (even) more effort or quickly ask for help with particular problems.
But shouldn’t we worry that this kind of information could simply kill their motivation? Are we not potentially driving students to drop a course, because they don’t know how to improve their apparently less successful learning behaviour? And what about those whose learning success is (significantly) above average? Will our data inadvertently undermine exceptional performance?
What happens to learning data?
What about the security and ownership of learning data? What are universities such as ETH doing to prevent such data being hacked? Future employers could thereby gain information on how consistently and quickly someone learned at university or where their individual strengths and weaknesses lie. And who does this learning data actually belong to? In principle, the data belongs to the individual students. But how can they control who is using their data and what they are using it for? Would individual data accounts be a solution? Or would such accounts cause additional problems, similar to the ones found with bank accounts, where the account holders frequently lack the basic skills needed to manage them appropriately?
Don’t avoid it, but...
The digitalisation of teaching – and the huge amounts of learning data this creates – has interesting potential, but it also opens up a broad range of unclear issues that we need to deal with. Should we therefore avoid producing this data? No, of course not – but if we want to use learning data successfully, we have to make efforts in other areas as well.
This involves more social science research and individual coaching for students in order to ensure that the learning data they produce offers a reliable and useful basis for lecturers and the students themselves. There also needs to be transparency, clarity and enforceability when it comes to data ownership. The students must be aware of their ownership rights and must be taught to manage their data skilfully. The next step has to be an ETH Zurich learning analytics strategy!
Comments
Additional comments are available for the German version of this article. Show all comments
Let me shortly elaborate on one of the many questions: Do students benefit from their generated learning data? One big challenge at ETH is that we give students very limited feedback during the first semesters. Only after getting the verdict from the base year exam do the students get a first answer to whether they're on track. Digital methods can give instant feedback to students and can keep track of their overall learning progress in a much better way than teaching assistants can, with their weekly exercise correction cycle. There is huge potential in digital methods. I think it should be normal to have an algorithm that suggests math problems to me that are exactly on the level of difficulty that is best suited for me to advance most effectively (consider applying the spaced repetition ideas that are by now well established in digital flash card apps, but for technical questions). But I'm not aware that such a tool is available, especially not on MOODLE or OLAT. With all the praise for MOODLE and OLAT, we should be aware that those platforms are extremely limited in their possibilities and are far from exploiting the full potential of digital learning methods. Thus it's important to ask these questions, but it's also important that we move on and then actually make use of the digital potential instead of using up all our energy on asking questions.
Dear Alexander, many thanks for your remarks. I agree that more feedback technologies should be available – but my point is that focusing on technology alone might not be sufficient. Students have to be enabled to adequately cope with their feedbacks and we should not be naïve about what happens with students’ learning data….