Teachers’ Affective Involvement in GenAI-supported Educational Expression: Exploratory Associations with GenAI Use Frequency, Task-specific Capability Beliefs, and Cultural-role Orientation

Abstract

Research on teachers’ generative artificial intelligence (GenAI) use has focused on adoption, readiness, and literacy, yet less is known about affective involvement in GenAI-supported educational expression among prior users. This cross-sectional study examined 149 teachers in a culturally diverse county in Southwest China. GenAI use frequency was examined alongside brief, study-specific indices of task-specific capability beliefs (TSCB), cultural-role orientation (CRO), and affective involvement. Polychoric parallel analysis and exploratory factor analysis provisionally separated the item sets; hierarchical regressions used HC3 robust standard errors. Use frequency correlated with TSCB (r = .356, p < .001) but was not independently associated with affective involvement in the full model (β = −.053, p = .519). TSCB scores accounted for the largest increase in explained variance (ΔR² = .196; β = .438, p < .001), whereas CRO added a smaller increment (ΔR² = .034; β = .199, p = .010). The full model explained 28.0% of the variance. Among prior users, weekly frequency alone provided a limited account of affective involvement. Professional learning should therefore assess what teachers can do in bounded educational tasks, how they justify their decisions, and whether that learning transfers to practice, rather than relying on use counts alone.