University teachers’ intention to integrate generative AI into teaching: The roles of GAI self-efficacy and GAI learning anxiety

Abstract

This dataset contains anonymised survey responses from 825 in-service Chinese university teachers, collected to examine how cognitive appraisals of generative AI (GAI) shape their intention to integrate GAI in teaching. The questionnaire operationalises seven constructs from an integrated UTAUT–Cognition–Affect–Conation framework: performance expectancy, effort expectancy, social influence, facilitating conditions, GAI self-efficacy, GAI learning anxiety, and integration intention, alongside demographic and career-stage variables. The data support the analyses reported in the associated manuscript, including covariance-based structural equation modelling (SEM), multi-group comparisons across career stages, and fuzzy-set qualitative comparative analysis (fsQCA).