Student reactions to AI versus human feedback in teamwork skills assessment
Igor Kotlyar, Joe Krasman
Abstract
Abstract As AI technologies become increasingly integrated into education, this research investigates how students react to AI-generated versus human feedback in teamwork skills assessment. In Study 1, 108 students completed a virtual teamwork simulation and received assessment feedback framed as either AI- or human-generated. Students showed a clear preference for human feedback over AI feedback, revealing a bias against machine-generated evaluations. Study 2, a scenario-based experiment involving 322 students, confirmed these findings and tested whether enhancing AI feedback with credibility and empathy cues could improve perceptions. These enhancements significantly improved reactions to AI feedback, and when both credibility and empathy were emphasized, reactions approached those for unenhanced human feedback. However, even with enhancements, AI feedback did not fully match the positive perceptions of human feedback. These findings highlight the need for thoughtful design to mitigate biases against AI feedback and suggest that blending AI and human feedback offers an effective approach for improving acceptance and engagement in educational contexts.