A feedback component that leverages counterfactual explanations for smart learning support
Sebastian A. Günther, Felix Haag, Konstantin Hopf, Philipp Handschuh, Maria Klose, Thorsten Staake · Perspektiven der Hochschuldidaktik · 2023
Abstract The growing prevalence of digital learning in higher education is accompanied by challenges regarding students’ self-regulated learning. While there is a plethora of behavioral interventions that aim at supporting students’ self-regulated learning, they often do not consider the heterogeneity of students in their intervention design. This paper presents a novel feedback intervention that leverages the potential of machine learning and counterfactual explanations for providing personalized feedback to support students’ learning. Ultimately, this approach could automatically adapt to different courses and thereby empower scalable and effective feedback.