Project IMPACT

• Duration: 2021 – 2025
• Project lead: Prof. Dr. Andreas Breiter, Prof. Dr. Karsten D. Wolf

Project management: Prof. Dr. Karsten D. Wolf & Prof. Dr. Andreas Breiter

Funding institution: BMBF – Federal Ministry of Education and Research

The joint project IMPACT promotes the improvement of higher education through the scalable use of artificial intelligence (AI) methods for the (partially) automated analysis of texts. The fundamental aim of the project is to enable new forms of sustainable support for information and learning processes for students and at the same time to relieve the burden on teaching staff in the supervision process. Along the student life cycle, prospective students, first-year students and students receive text-based, highly informative and personalized feedback in the orientation and entry phase, during the course of studies (formative assessment) and at the end of their studies (summative assessment). This not only helps to cope with paracurricular study requirements, promote individual learning goals and self-regulation strategies for future learning or reduce uncertainty and excessive demands on students, but also supports teachers and promotes active engagement with AI-generated feedback by students and teachers.

At five German universities (Goethe University Frankfurt, Humboldt University Berlin, Freie Universität Berlin, Fernuniversität Hagen and the University of Bremen), text-based AI processes such as chatbots and personalized feedback systems for formative and summative assessment are being introduced into widespread use, accompanied by change management based on the SHEILA process model. The interdisciplinary consortium uses internationally proven open source software solutions (Rasa X, OnTask), common standards and interoperability in university teaching with learning management systems (Moodle, Stud.IP, ILIAS), online study choice assistants and assessment systems (LPLUS). (Media) didactic concepts and adaptations are made for the targeted integration of AI applications.

For more than 10 years, the research field of learning analytics has focused primarily on the application of AI in higher education. The IMPACT joint project brings together five universities with particular expertise in the field of ethical, legal and social implications (ELSI) and a wealth of practical experience in the didactic use of learning analytics. In piloting and evaluation studies, requirements arising from the diversity of the target groups (e.g. gender-related differences) are taken into account. There is no question that the project and learning analytics in general will be implemented in Germany in accordance with the applicable European Data Protection Directive, the GDPR and the corresponding state data protection laws. The partner universities therefore speak of “Trusted Learning Analytics” (TLA), which is fundamentally different from other learning analytics approaches. TLA provides for the ethically responsible, targeted and transparent use of data, which can give German universities a unique selling point in international comparison.

A requirement for the sustainable implementation of AI applications in studies and teaching is data-ethical change management, which is developed and applied in all participating universities in order to enable continuous cooperation with committees, teaching staff and students. The project results will be made accessible throughout Germany through workshops and according to open science principles by the network.