BeLEARN, AutoGrade4Gym

AutoGrade4Gym

A tool for the automatic correction of Python programmes in upper secondary education.

Duration: July 2023 – June 2024
Status: Completed
Educational Level: Upper Secondary Level – Grammar School Education
Topics: Digital Skills & Literacy, Digital Tools
Keywords: Digital Skills, Learning Platforms

Initial Situation

Computer science was introduced as a compulsory subject in the Federal Maturity Recognition Regulation (MAV) in 2018. With the complete revision of the MAV, adopted by the Federal Council on 28 June 2023, computer science has been upgraded to a core subject and now counts towards passing the Matura. Initial experiences at grammar schools show that, particularly in programming education, there is not enough opportunity for practice given the current resources.

Objectives

The goal of this project is to develop a tool that allows programming tasks to be automatically checked and corrected. In addition, a comprehensive collection of tasks (including test data) is being created, from which teachers can select suitable exercises. Students can use the tool to learn independently, receive immediate feedback and are motivated through gamification to continue making progress. Teachers no longer need to correct every solution but can provide targeted and efficient support.

Method

We achieved the following milestones:

  • Spring 2023: First prototype created as part of the “Software Engineering Lab” course at the University of Bern
  • 23.08.2023: Project presentation and introduction to grammar school teachers
  • Autumn 2023: Implementation of a new version by Michael Kaiser
  • Spring 2024: Creation of a task collection
  • Spring 2024: Testing phase with grammar school classes

Results

In general, the grader tool was considered very useful. Automated programming tasks provide effective support for teaching and allow students to practise programming skills independently. The gamification features (especially feedback in the form of memes) were perceived as a motivating enhancement. However, the feedback given by the grader still needs to be more precise. There were also some technical issues, such as sorting tasks or website refreshes not executing automatically. Usability also requires further improvement.

Implemented Translation

The current state of development is that of a first, minimally functional iteration. The grader can be used productively and provides real value. We received initial feedback from Gymnasium Lerbermatt and Gymnasium Kirchenfeld, which is now being used to develop the next, improved version. The next version will provide improved feedback on task corrections. In addition, a feature will be implemented that allows students to request hints for tasks. Implementing these enhancements is essential for a successful and broader use of the grader at grammar schools.

Publications

Final Report (in German)

Project Lead

BeLEARN, AutoGrade4Gym
Prof. Dr. Thomas Studer Institute for Computer Science, University of Bern

Project Collaborators

BeLEARN, AutoGrade4Gym
Prof. Dr. Martin Lehmann Institute for Upper Secondary Level, PHBern

Participating Institutions