Information and Computing Sciences Colloquium

Automated Feedback for Learning Programming

Hieke Keuning

Date: 16:00 – 16:30, Thursday, 28.01.2021
Location: Gather Town

Title: Automated Feedback for Learning Programming
Abstract: Learning how to program has been studied ever since the first programmers were trained in the 60s. From the very beginning of this research field, tools have been developed to support students in their learning. A central element in these tools is feedback, which has the potential to greatly affect learning, if delivered properly. Feedback on programming exercises could be in the form of giving hints on how to correct errors, showing which test cases fail, or calculating a grade. If the feedback is automated, it alleviates teachers from the effort of giving feedback to large groups of students.

In this talk I will give an overview of the types of feedback these systems present to students, and the techniques used to generate it. We have found that feedback mostly focuses on identifying mistakes and less on fixing problems, taking a next step, and writing code with good style and quality. In the second part of this talk I will elaborate on a tutoring system we designed that helps students to improve their code, focusing on the small programs they write as beginners. I will discuss its motivation, design and how a large group of CS students worked with the system.