For Better Feedback in Online Teaching
Currently, almost all students learn only digitally at home, because in the wake of the Covid 19 pandemic, universities have largely converted their teaching to online courses. But virtual lectures and seminars often still lack individual feedback for learners. A new project being conducted by the DIPF | Leibniz Institute for Human Development and Educational Information and Goethe University Frankfurt aims to change that. The researchers want to develop a software tool that uses modern, automated evaluation methods to help provide accurate feedback.
"Online teaching, which is predominantly carried out at the moment, also offers many opportunities. For example, large amounts of data on learning progress and learning outcomes are generated. With advanced analysis tools, this data can be used to precisely address the needs of students - in a way that is conducive to learning and at the same time compliant with data protection laws," explains Prof. Dr. Hendrik Drachsler from the "Educational Technologies" department at the DIPF. He is leading the "HIKOF-DL" project together with Prof. Dr. Andreas Frey from the "Educational Psychology" department at Goethe University and with Prof. Dr. Alexander Tillmann, Managing Director of "studiumdigitale", the central e-learning facility at Goethe University. The project, which is scheduled to run for three years and has now been launched, is being funded with around 650,000 euros from the Distr@l funding program of the state of Hesse. The program supports innovative, application-oriented projects in the field of digitization.
Learning analytics and digitally supported competence diagnostics
The project team builds on the expertise and developments that have already taken place in two specialist areas: 1. learning analytics and 2. digitally supported competence diagnostics. The goal of learning analytics is to analyze data from teachers and learners in near real-time to optimize teaching-learning processes. Psychometric competence diagnostics, in turn, makes it possible to record the individually achieved learning status of pupils or students systematically, precisely and with direct reference to action. "For learning analytics as well as for competence diagnostics, artificial intelligence methods are increasingly being used to make data collection and evaluation more individualized and automated. We want to build on this," says Professor Drachsler. On this basis, a new, versatile software application is to be developed in the course of the project. It will enable students to receive targeted feedback during the learning process or after learning units - even in large lectures with many participants. The solution will have an open programming code (open source) and will strictly comply with the European data protection regulation.
The tool will be able to detect at an early stage that certain learners are not progressing well in a topic - for example, by comparing the pace of work with that of previous classes. The software will inform the lecturers about this and at the same time give hints as to what the problem might be. Perhaps not all research possibilities have been used or the exchange with fellow students is missing. The lecturers could then recommend further materials or the participation in an online working group to the students. The IT solution will also enable final exams that adapt to the individual performance of the students. The resulting data will also be evaluated for the ongoing learning process. In addition to quantitative grading by the seminar instructor, the tool will provide students with an assessment of their individual level of competence. This automated and written feedback will include an analysis of the learners' strengths and weaknesses as well as recommendations for future learning and the appropriate tools for this.
The project also aims to create the conditions for the IT solution to be transferred into practice in the best possible way. For example, the finished software is to be evaluated in the context of the largest possible lecture at Goethe University with more than 1,000 participants. On the one hand, the universities are to profit from the work. On the other hand, companies and private education providers should also take up the results. This knowledge transfer will be promoted by a separate project advisory board. It is composed of representatives of start-ups and larger companies, who will review the developments of HIKOF-DL twice a year to determine their applicability in the business sector.