Use hybrid human–AI feedback models that combine AI's and instructors' complementary strengths¶

Strategy · All strategies
Evidence · 6 claims (5 for, 1 against) · 4 studies (1 causal, 1 associational, 1 qualitative, 1 design), q2–q3 · 0 of 4 report an effect size · 6 claims rest on one study

Description¶

The authors propose hybrid feedback models enabling human–AI collaboration in feedback provision, described as "a promising avenue for a scalable yet effective feedback practice." Two approaches are outlined: AI provides an initial round of feedback that instructors then review and enrich, or AI handles specific aspects such as syntax errors and formatting while instructors provide higher-order feedback on program logic and code efficiency.

Design Implications¶

Context¶

Requirements¶

• Instructor capacity to review or complement AI-generated feedback; clear division of feedback responsibilities between AI and instructor

Constraints¶

• Presented as a design recommendation; the article did not test a hybrid model empirically

Target Learners¶

• higher education students in large programming courses

Target Learning Goals¶

• scalable and effective feedback on programming tasks

Claims¶

• First-year EAP students prefer a hybrid feedback model combining AI-generated comments with instructor feedback over either source alone [+W]
• Business Writing students valued AI feedback for higher-order and sentence-level comments but found it sometimes too general, while appreciating peers' contextual knowledge [+W]
• Students bound their reliance on AI feedback and advocate a hybrid model assigning AI broad language concerns and instructors individualized relational guidance [+W]
• Both instructor and AI-generated feedback were perceived positively, with instructor feedback rated higher descriptively on all four perception dimensions [+W]
• Students receiving instructor feedback showed significantly greater improvements in lab assignment scores than students receiving AI-generated feedback, after controlling for initial knowledge [-W]
• Students perceived instructor feedback as significantly more useful than AI-generated feedback, with no significant differences on fairness, developmental value or encouragement [+W]

Related Strategies¶

• Combine GAI tools with traditional methods in a hybrid approach to balance efficient collaboration and critical evaluation
• Train AI feedback models on context-specific educational data including student, course and instructor data
• Integrate ChatGPT as a feedback resource alongside peer feedback for essay writing in higher education

Examples¶

-

Key Sources¶

• Er, Erkan; Akçapınar, Gökhan; Bayazıt, Alper; Noroozi, Omid; Banihashem, Seyyed Kazem. (2024). Assessing student perceptions and use of instructor versus AI-generated feedback. British Journal of Educational Technology. https://doi.org/10.1111/bjet.13558