AI Assisted Fair and Efficient Grading in the Statistics and Data Science Classrooms


Mine Dogucu (University of California Irvine)


Abstract

Assessments are a key component of learning. However, providing feedback on assessments (with or without points assigned) is time consuming for instructors and graders. This becomes a bigger challenge in statistics and data science classrooms as the class enrollments continue to increase. This talk will introduce Gradescope, an online platform for grading which is designed to provide feedback efficiently. Experiences from using Gradescope in classrooms with varying sizes, from introductory level to more advanced levels will be shared. Different assessment types including handwritten homework assignments, multiple choice tests, coding assignments, and open-ended questions from online and on paper assignments and how to grade these assignments efficiently will be shared. Technical features of Gradescope, including AI assisted grading will be shared and its shortcomings in the current version will be noted.


Recording