Diving into Software Evolution: Virtual Reality vs. On-Screen

Abstract

Abstract—Background: Traditional 2D visualizations have been widely used for software metrics and evolution analysis, offering structured views of complex data. However, the advent of Virtual Reality (VR) technologies introduces new possibilities for immersive and interactive software visualization, potentially enhancing comprehension and user engagement. Objective/Aim: This report aims to evaluate the effectiveness of VR immersive visualizations compared to traditional 2D on- screen visualizations for understanding code metrics across soft- ware releases. Specifically, we seek to determine if VR provides better accuracy and speed for comprehending high-level software evolution tasks. Method: We will conduct a controlled experiment with 30 participants from academia and industry, using GitLab, GitHub, or an IDE for on-screen visualizations and a VR setup using Meta Quest 3. Tasks related to software evolution will be completed in both settings. Accuracy and time will be measured and analyzed using mixed linear models and non-parametric tests to compare the two approaches. Data will be sourced from GitHub repositories with similar project characteristics.