Growing grade-by-grade

Abstract

The enrollment of computer science students continues to increase, with record enrollment numbers in the course ”Data Structures and Algorithms” in fall 2022. As a result, teaching methods and systems must evolve to support student progress in the face of scarce teaching resources. This paper examines the shift from manual to peer-reviewed and auto-graded grading processes and investigates students’ perceptions of different grading styles. The developed automatic graders utilize data from both the Plussa LMS and GitLab, which serves as the channel for student submissions and provides feedback for formative assessments. The results indicate that peer-reviews are accepted as an exercise for the reviewer, but not as a means to grade the reviewee. Autograders are well-received due to their instant feedback, ability to allow for multiple submissions and iterate towards more efficient solutions, which helps foster a growth mindset.