Exploring the Relationship between Human-AI Interaction Experiences and Teacher' Intention to Use AI in Teaching: Evidence from Stem and Non-Stem Teachers in Vietnam

Exploring the Relationship between Human-AI Interaction Experiences and Teacher' Intention to Use AI in Teaching: Evidence from Stem and Non-Stem Teachers in Vietnam

This study explores how human-AI interaction experiences developed through AI professional development (PD) relate to teachers' intention to use AI in teaching, with particular attention to differences between STEM and non-STEM teachers in Vietnam. Grounded in Expectancy-Value Theory, the study proposes a structural model in which Human-AI interaction experiences are linked to AI self-efficacy, professional value, and trust, which in turn are associated with teachers' intention to use AI in teaching. Data were collected from 97 secondary school teachers who completed a two-day AI-focused PD program and were analyzed using PLS-SEM with multi-group analysis. Results indicate that human-AI interaction experiences were significantly and positively associated with AI self-efficacy, professional value, and trust. Among the proximal predictors, AI self-efficacy showed the strongest association with intention, followed by professional value, while trust did not show a significant direct association in the full sample. Multi-group findings suggest disciplinary differences: AI self-efficacy was a stronger predictor among STEM teachers, whereas professional value and trust played comparatively stronger roles among non-STEM teachers. Based on a cross-sectional post-intervention design, the study provides preliminary evidence on the role of experiential AI training and suggests the potential value of differentiated PD strategies to support AI integration across subject areas.