Differential Effects of AI Anthropomorphism on Learning: An Experimental Investigation
Abstract
This study examines how two key anthropomorphic features of AI teaching agents—agency (proactive scaffolding) and sociability (social cues)—differentially affect learning outcomes (retention vs. transfer) through affective, cognitive, and trust-based mediating pathways. We employ a 2×2 between-subjects experimental design with 200 participants learning about virus infection processes. This preregistration documents our hypotheses, design, and analysis plan prior to data collection.
Fields of Study & Subjects
Intelligent Tutoring Systems and Adaptive Learning
Social Robot Interaction and HRI
Artificial Intelligence in Education
Publication details
- Journal / source
- OSF Preprints (OSF Preprints)
- Publication date
- 10 March 2026
- Language
- EN
- Document type
- Preprint