ReflectBuddy: Where AI Supports Authentic Reflective Learning
Date: Wednesday 11 June 2025
ReflectBuddy recently won the 2025 Brinksma Innovation Grant (BIG) for its AI-assisted feedback tool aimed at improving reflective writing and streamlining formative assessment within the Chemical Science & Engineering bachelor’s programme at the University of Twente. Selected from a competitive group, ReflectBuddy caught attention for its practical approach to scaling reflective learning.
Meet the Team Behind ReflectBuddy
Leonie leads the Chemical Science & Engineering programme and established its essential-skills learning line. She and her colleagues developed workshops to foster student skill development. With the third cohort engaged, timely feedback became a challenge, inspiring the ReflectBuddy concept.
Linlin is an educational specialist who co-developed the skills workshops. Her role includes shaping the proposal and focusing on evidence-based research to support and evaluate ReflectBuddy.
Jéré teaches essential skills in the Bachelor’s and supports the Master’s programme. He contributed to project conceptualization and document reviews.
Arnoud develops workshops and provides feedback on student reflections each quartile. His experience with local language models helped optimize the proposal’s feasibility and alignment with educational practice.
Why Focus on Reflection?
ReflectBuddy centers on AI-supported feedback for reflective writing rather than other AI applications like grading or adaptive learning.
Leonie and Linlin highlight that skill development depends heavily on reflection, citing Dewey: “We do not learn from experience … we learn from reflecting on experience.” Yet, students often struggle to go beyond surface-level reflections, and providing timely, quality feedback is demanding for educators. AI offers a way to assist both students and teachers in this process.
Jéré adds that current AI excels with text-based tasks but is less effective for subjects like math. Using AI as a language algorithm suits reflective writing better than technical grading.
The Hybrid Feedback Model
ReflectBuddy combines AI-generated feedback with teacher refinement instead of fully automating feedback.
Linlin explains that AI can offer a first layer of feedback, such as distinguishing descriptive from analytical content. However, teachers remain vital for interpreting nuances and delivering personalized guidance. Refining AI feedback also ensures accuracy and builds student trust.
Arnoud notes this model allows for iterative improvements. If AI feedback shows recurring issues, prompting strategies can be adjusted quickly.
Jéré emphasizes the personal nature of reflections. Students expect a human to read their work, and AI alone can’t be fully trusted for factual accuracy.
Leonie points out that expanding lecturer access risks breaching student privacy. ReflectBuddy offers a secure digital space where a small team can enhance feedback quality and share it promptly, making it more effective.
What Made ReflectBuddy Stand Out?
Several strong proposals competed for the BIG grant. The team believes their focus on reducing teacher workload and streamlining feedback was particularly relevant in today’s educational climate.
Linlin acknowledges the quality of other projects but notes limited insight into their full proposals.
Jéré feels their straightforward approach to easing educators’ tasks resonated with the jury.
Looking Ahead: Potential and Support
While ReflectBuddy currently targets reflective writing, the team sees potential for its approach to expand to other assignments and programmes.
Linlin mentions further development is needed before sharing concrete plans.
Jéré stresses the importance of improving AI literacy among educators. Understanding what AI can and cannot do is crucial for its effective integration into teaching.
Contributing to Educational Innovation
Linlin says ReflectBuddy complements a wider skills development initiative relevant to many higher education programmes. They welcome collaboration with others facing similar challenges.
Jéré points out that AI is unavoidable in education. This project helps explore practical uses of AI for formative assessment.
Arnoud emphasizes testing innovations in real educational settings. The project includes research comparing AI-assisted feedback with traditional methods and measures time saved. These results can guide other educators considering AI integration.
ReflectBuddy demonstrates how AI can assist education without losing the human element, offering a promising model for AI-supported learning and assessment.
Your membership also unlocks:
