Birte Heinemann

 

How can different domains and areas be improved through the use of multimodal learning analytics?

Learning analytics offers the opportunity to gain a unique view into the cognitive processes of learning. Technological developments in recent years allow us to explore learning from multiple perspectives, better understand the process, and provide increasingly sophisticated feedback and assistance for students.
Various sensors can give insights into the complex processes that have an impact on the learning process. But we need researchers with ideas and knowledge from different perspectives to draw insights from learning data, to improve learning units and feedback. And extending research structures to include multimodal learning analytics is still a challenge.

Part of this dissertation is to look at the research process in the interdisciplinary field of learning technologies and build on that to develop structures that lead to sustainable open projects with comparable data to better connect learning analytics research with subject area experts and didactics.
For this purpose, I am testing the structures developed in the DigiLab4U research project, in which we are developing learning analytics for lab-based learning. The results from the infrastructure developments contribute especially to the practical part of the dissertation, which focuses on research on concrete learning applications. Here, one subarea is to support the investigation and evaluation of learning applications with multimodal learning analytics.
In various use cases, the question of how MMLA can be used in learning scenarios and what conclusions can be drawn for the development of learning technologies is explored. This part focuses on learning applications in virtual reality, as this technology provides good conditions for multimodal learning analytics. A learning scenario is taken from the core topics of computer science and is planned and developed in an agile way using didactic methods. Here, the focus is on multimodal data collection in addition to didactic aspects of learning and teaching in VR. An important reason for the decision to focus on the basics of the rendering pipeline in the learning environment is the incentive to use the possibilities of virtual reality as efficiently as possible. The mixture of VR as technology and this topic can interactively contribute to the learners' understanding on different levels. The learning objectives of this application are in various areas and include action-oriented knowledge as well as cognitive knowledge.

Furthermore, in this research, existing learning scenarios are equipped with the possibility to collect multimodal learning data, which also integrates them better in interdisciplinary research processes. The data collected will open up new insights and (hopefully/possibly) enable transferable knowledge to sustainably improve learning in VR and beyond in various learning scenarios.