Supporting learning by means of learning analytics

Abstract

In the Object-Oriented Programming course (TDT4100) at NTNU, a lot of data is being collected about how the students work on programming assignments. The data is collected through a Learning analytics extension for the Eclipse Integrated development environment (IDE). This extension displays the collected data back to the students, with the goal of stimulating reflection and self-evaluation regarding how they work on programming assignments. In this project, we explore the collected data using Learning analytics approaches. Through an iterative process, we conduct five experiments with the goal of identifying distinct programming modes. We identify the "struggling" mode and design algorithms to automatically detect this mode. In parallel to the experiments, we design and implement a tool for visualization and analysis, which will streamline our experiment process.