Workload perception in educational resource recommendation supported by artificial intelligence: a controlled experiment with teachers

Abstract

Researchers are increasingly interested in enabling teachers to monitor and adapt gamification design in the context of intelligent tutoring systems (ITSs). These contributions rely on teachers’ needs and preferences to adjust the gamification design according to student performance. This work extends previous studies on teachers’ perception of their cognitive effort and dedication to creating and monitoring educational resource recommendations on a simulated gamified educational platform. This study compares teachers’ perceptions of workload using one of three scenarios— manual, automated, and semi-automated—to recommend educational resources through a randomized experiment. In this study, 151 participating teachers evaluated their perception of cognitive effort and time dedicated to creating recommendations for missions and monitoring students on the platform. The results indicate that the teachers’ perception that the automated scenario has a lower workload than the manual scenario significantly raises the hypotheses. Our results also suggest that teachers’ perception of the textbook scenario is different according to their level of knowledge about Information and Communication Technologies (ICT). For teachers with advanced ICT knowledge, the manual scenario is perceived as a scenario that indicates a more outstanding performance. According to the educational level of the teachers, the perception of mental demand for the automated scenario is significantly different. These significantly contribute to understanding teachers’ perceptions when using educational platforms in their classes.