The Design of an Adaptive Tool Supporting Formative Assessment in Data Science Courses.

The automated correction of assignments is a task that received large support from artificial intelligence. In previous research, we approached the problem of automatically providing a grade and feedback to students in solving data science exercises, that resulted in the development of the rDSA tool. In this paper, we discuss the first steps towards the development of an adaptive system – based on the rDSA tool – supporting the students’ formative assessment activities. In particular, we present the context of use, the requirements – elicited through a study with a small cohort of students, the models enabling adaptation, and the user interface. Finally, we evaluated the user interface through a further study that involved both qualitative and quantitative measures.