Giving Voice Within a Voice: Rethinking Feedback and Assessment with GenAI

• In Event: AERA Roundtable Session Wednesday 11:45 am - Gold 3
  In Roundtable Session: AI in Practice: Adoption and Reflection

Abstract

This study explores how generative AI (GenAI) acts as a “second voice” in formative assessment, transforming feedback into a dialogic, co-constructed process between instructors, students, and technology. Through a mixed-methods study across Canadian and Chinese universities, GenAI tools supported learners’ in English writing, with and without GenAI. Survey (N=207) and interview (N=32) data revealed that GenAI fostered metacognitive awareness and student agency, reframing feedback as iterative and reflective. Instructors noted a shift toward shared co-agency, though concerns about quality and overreliance persisted. The metaphor “giving voice within a voice” captures this dual dialogue, highlighting GenAI’s potential to democratize assessment while underscoring ethical challenges. Findings advocate for human-AI collaboration to reimagine feedback as a constructivist, equitable practice.

Authors

• Nadia Delanoy, University of Calgary

• Danni Chen, Harbin Institute of Technology