What is it about?
This study investigated how the organization and summarization of multi-peer feedback affect students' ability to review and implement it in software engineering education. Recognizing that receiving diverse feedback from multiple peers can be overwhelming, the research aimed to determine if AI-based automatic summarization could streamline this process. The study utilized an exploratory focus group methodology with master's students, who evaluated three distinct AI-generated summary designs: categorized, textual, and list-based, created using ChatGPT-4 Turbo from real peer feedback. The core research questions focused on how summarizing similar and contradictory comments and different visual organizations influenced the feedback review process and the likelihood of implementation. The findings highlighted that both the content and the visual organization of summarized feedback significantly contribute to its implementation likelihood.
Featured Image
Why is it important?
This study is important because it addresses the challenge of overwhelming multi-peer feedback in software engineering education. While peer assessment is valuable for learning and simulating industry practices, the sheer volume and diversity of feedback can be time-consuming and demotivating for students, hindering their ability to review and implement it effectively. The research demonstrates that AI-based automatic summarization, particularly with effective visual organization, can significantly alleviate this burden, making feedback more manageable and increasing its impact on student learning.
Perspectives
I was motivated by the belief that AI-based automatic summarization, combined with clear visual organization, can offer a practical way to help students make sense of diverse feedback and turn it into meaningful improvements. I hope these findings will inspire the development of more actionable feedback systems. While this was an exploratory study with a limited scope, it is a valuable first step. I am committed to continuing this research, gathering more robust, quantitative data to refine these approaches and expand their applicability to broader and more varied contexts.
Somayeh Bayat Esfandani
Norwegian University of Science and Technology
Read the Original
This page is a summary of: The Impact of Multi-Peer Feedback Summary Organization on Review and Implementation of Feedback, June 2025, ACM (Association for Computing Machinery),
DOI: 10.1145/3696630.3727253.
You can read the full text:
Contributors
The following have contributed to this page
