Hybrid intelligent feedback, which combines human judgment with generative AI capabilities, offers a promising approach to enhance educational feedback by balancing technological efficiency with human contextual understanding and empathy.

Objective: The main goal of this study was to propose a pedagogical framework for hybrid intelligent feedback that integrates generative AI (GenAI) with human feedback processes, addressing the lack of strong pedagogical foundations for implementing GenAI in educational feedback systems.

Methods: The authors developed a conceptual framework through analysis of existing literature on feedback theory, artificial intelligence in education, and hybrid intelligence. They explored the distinctions between human and artificial cognition, positioned GenAI feedback within established feedback theory, and proposed a structured approach for implementing hybrid intelligent feedback in educational settings. The framework was constructed by examining traditional human-centered feedback challenges and identifying ways that GenAI could complement human feedback providers.

Key Findings:

• Traditional human-centered feedback (from teachers, peers, and learners themselves) faces challenges including time constraints, expertise limitations, and self-assessment difficulties.
• Human and artificial cognition differ fundamentally: human cognition involves emotional awareness and contextual understanding, while AI excels at pattern recognition and data processing but lacks adaptability and causal reasoning.
• GenAI feedback can serve two main roles: as an independent feedback source or in collaboration with humans (hybrid intelligent feedback).
• The paper identifies three approaches to hybrid intelligent feedback: Human-Led Feedback with GenAI Support, Adaptive Human-GenAI Feedback, and GenAI-Led Feedback with Human Enrichment.
• Effective hybrid intelligent feedback follows five key principles: complementarity, iterative refinement, dynamic adaptability, enhanced personalization, and shared agency.
• The proposed pedagogical framework outlines a six-step implementation process: determine feedback purpose, align with student stages, decide on hybrid-intelligent approach, generate feedback, evaluate feedback, and implement feedback.

Implications:

• The framework provides educators with a structured approach to integrate GenAI into feedback processes while maintaining pedagogical integrity.
• Hybrid intelligent feedback can potentially overcome limitations of traditional feedback sources by combining AI efficiency with human contextual understanding.
• The complementary strengths of humans and AI can enhance the quality, timeliness, and personalization of feedback.
• The framework supports a balanced application of human judgment and AI efficiency, avoiding both technological determinism and the neglect of human-centered education.
• By clarifying the roles and relationships between human and AI feedback providers, the framework helps establish trust and meaningful engagement with GenAI feedback.

Limitations:

• The framework is primarily conceptual and requires empirical validation in various educational contexts.
• The effectiveness of hybrid intelligent feedback may vary depending on the specific technologies, subject areas, and student populations involved.
• Ethical concerns regarding transparency, accountability, and privacy need further exploration.
• The framework does not fully address how to develop GenAI feedback literacy among students and teachers.
• Implementation challenges, including technical infrastructure and teacher training, are not extensively discussed.

Future Directions:

• Develop detailed pedagogical guidelines for implementing hybrid intelligent feedback in various educational contexts.
• Enhance methodological approaches to improve GenAI's ability to provide contextually relevant feedback.
• Explore different models of human-GenAI collaboration in feedback processes.
• Investigate how to develop hybrid intelligent feedback literacy among students and teachers.
• Research how hybrid intelligent feedback can support self-regulated learning and metacognitive awareness.
• Address ethical considerations around transparency, accountability, and trust in hybrid intelligent feedback systems.

Title and Authors: "Pedagogical Framework for Hybrid Intelligent Feedback" by Seyyed Kazem Banihashem, Omid Noroozi, Hassan Khosravi, Christian D. Schunn, and Hendrik Drachsler.

Published On: April 29, 2025

Published By: Innovations in Education and Teaching International (Taylor & Francis Group)