Paper Type
Short
Paper Number
PACIS2026-1715
Description
This paper examines how AI-powered tutor design shapes students’ learning experiences, focusing on response quality and anthropomorphic cues. Drawing on the Stimulus-Organism-Response (S-O-R) framework, we conceptualize these design features as stimuli that influence learners’ internal states, including perceived usefulness, flow experience, psychological empowerment, and parasocial connection, as well as subsequent learning-related outcomes and behavioral intentions. We report two completed pilot studies and one ongoing pilot with university students that refine the experimental design, measures, and manipulation of tutor characteristics. Preliminary findings suggest that response quality plays a central role in students’ perceptions of AI tutors and their willingness to continue using AI-supported learning tools, while the value of anthropomorphic cues requires further controlled testing. The study contributes to AI tutor research by linking tutor design features to learner states and offers practical guidance for designing effective AI-mediated educational systems.
Recommended Citation
Guo, Yi Maggie and Syed, Urooj, "Designing AI Tutors: Response Quality, Anthropomorphism, and Student Learning" (2026). PACIS 2026 Proceedings. 8.
https://aisel.aisnet.org/pacis2026/is_education/is_education/8
Designing AI Tutors: Response Quality, Anthropomorphism, and Student Learning
This paper examines how AI-powered tutor design shapes students’ learning experiences, focusing on response quality and anthropomorphic cues. Drawing on the Stimulus-Organism-Response (S-O-R) framework, we conceptualize these design features as stimuli that influence learners’ internal states, including perceived usefulness, flow experience, psychological empowerment, and parasocial connection, as well as subsequent learning-related outcomes and behavioral intentions. We report two completed pilot studies and one ongoing pilot with university students that refine the experimental design, measures, and manipulation of tutor characteristics. Preliminary findings suggest that response quality plays a central role in students’ perceptions of AI tutors and their willingness to continue using AI-supported learning tools, while the value of anthropomorphic cues requires further controlled testing. The study contributes to AI tutor research by linking tutor design features to learner states and offers practical guidance for designing effective AI-mediated educational systems.
Comments
04-DigitalLearning