3.14 Case Study: Learning Analytics
Learning Analytics (LA) is discussed quite a lot in the context of the AI education discussion. The idea behind LA is very basically that the learning process or an educational offer should be optimized. To do this, data is collected and analyzed to identify potential problem areas and predict future learning performance.
A rough distinction can be made between three target levels:
- Descriptive analytics: What happens? How and when are e.g. online courses, formative assessments used? Can patterns in learning behaviour be identified?
- Predictive analytics: What will happen? Which learner groups are likely to form (e.g. particularly strong learners get together)? Which learners are at risk of dropping out?
- Prescriptive analytics: What can we do? How should learner groups be mixed? Which learning pathways should be provided for which groups of learners?
We talked with Prof. Dr. Sigrid Hartong from the Helmut-Schmidt University in Hamburg (Germany) about Learning Analytics. Her research and work focus on educational policy and governance, datafication and digitalization in educational institutions, building international research networks, and Critical Data Studies. Prof. Dr. Sigrid Hartong is part of the initiative UNBLACK THE BOX. It is a network initiative founded in 2019 by researchers from education science, sociology, information technology, media, and health education, as well as teachers in schools, universities, and pedagogical training. Their goal is to enable educational institutions and teachers to respond to the growing datafication and digitization of education with enlightened, critical, and conscious decision-making, even without extensive IT knowledge.
Here are some of the goals and associated opportunities of learning analytics:
– Prediction of expected learning outcomes, such as learning progress or learning results.
– Providing feedback and recommendations for relevant next learning steps and learning materials
– Personalization of learning environments
– Promoting reflexion and awareness of teaching-learning processes
However, LA also poses many risks and questions. For example, the social context is not taken into account in simple LA procedures. But is this not also relevant? Which data may be collected, and which are relevant at all? Relevance also depends on what is understood as learning success in the first place.
What is learning success for you, or what distinguishes it? Share your thoughts on this in the comments.
Website: UNBLACK THE BOX