AI vs. human: who makes students feel more? A preliminary comparison study of emotional engagement in peer review

As AI tools become increasingly integrated into educational feedback, understanding how AI and human reviewers shape students’ emotional experiences is important for effective feedback design. This study examines college students’ emotional engagement when receiving AI and human peer feedback across two undergraduate course cohorts at a US public research university. Reflections were analyzed grounded in Control-Value Theory (CVT) and Affective Events Theory (AET). CVT explains how different feedback sources trigger distinct appraisal processes, whereas AET explains how these appraisals accumulate into emotional patterns over repeated interactions. Findings reveal that AI peer review generated stable but predominantly neutral emotional states and sustained cognitive engagement, while human peer review motivated richer emotional experiences. Cross-cohort thematic analysis showed that reflections toward AI reviewer were dominated by Tool Mastery, whereas reflections regarding human reviewer shifted from Social Connection toward Empathy and Inconsistent Quality as the semester progressed. These findings suggest that AI and human peer review engage students through different emotional mechanisms. Extending CVT and AET to AI-mediated feedback, this study shows how appraisal differences develop into distinct emotional trajectories over time. The findings support hybrid feedback models that use AI to complement rather than replace human interaction, preserving both reliability and emotional engagement.