Same Feedback, Different Source: How AI vs. Human Feedback Shapes Learner Engagement
AI feedback is identical, but students still prefer humans. Here's why it matters...
Deep Dive
A new study reveals a major bias in education: learners who believe feedback comes from a human teaching assistant (TA) engage significantly more, despite receiving the exact same LLM-generated feedback as the AI group. The TA-attributed group showed effect sizes (d) from 0.88 to 1.56 for time and effort spent. Perceptions also differed sharply, with AI feedback ratings tied to prior trust (r=0.85) and human feedback to perceived genuineness (r=0.65).
Why It Matters
This 'source bias' challenges the effectiveness of hybrid AI-human educational systems and has major implications for how feedback is presented to learners.