Patrick Totzke is a professor in the Department of Computer Science at the University of Liverpool, where he teaches an introductory course in object-oriented programming. With around 400 first-year students each year, managing assignments and grading at scale has always been a priority. For the past four years, Patrick has relied on CodeGrade to simplify grading, enhance feedback, and create a more engaging learning experience for students.
Challenges Before CodeGrade
Patrick’s course transitions students from introductory Python or Haskell to Java. With such a large class size, grading assignments manually was overwhelming, even with previous autograding solutions. He also faced a growing challenge: designing assignments that couldn’t be easily solved using generative AI. While assignments are currently take-home, he’s considering shifting to shorter, in-person coding assignments to enhance learning. The usual take-home exercises will still be used, although their weight will be decreased.
Before CodeGrade, Patrick spent significant time managing grading logistics and troubleshooting submission issues. However, once CodeGrade was introduced, the process became much more streamlined. Everything is now graded automatically, giving students immediate feedback and allowing instructors to focus on higher-value feedback—reviewing code, adding inline comments, and providing meaningful insights beyond the automated assessments.
A New Approach to Feedback
One of the biggest transformations CodeGrade brought to Patrick’s course was in feedback delivery. The autograder handles the bulk of grading, ensuring fairness and consistency, while instructors use the freed-up time to focus on deeper engagement with student work.
“We are not the bad cops... we are helping them across the line,” Patrick explained. Instead of spending all their time marking assignments, instructors now guide students toward better coding practices and deeper understanding.
This structured feedback approach has led to a noticeable shift in student engagement. Instead of waiting days for results, students can quickly iterate on their work, learning from mistakes in real time.