Institutional digitalization and digital stress: a four-wave observational study of teacher-reported AI-collaborative teaching in physical education

How institutional digitalization relates to teachers' experience of work and subsequent artificial intelligence (AI)-supported practice remains unclear.

Drawing on job demands-resources principles, we examined these relationships in a four-wave observational cohort of 961 physical education teachers from 120 Chinese institutions. We linked an organizational-source digitalization proxy to teacher questionnaires and a study-specific teacher-reported score of AI involvement in teaching tasks. Analyses adjusted for baseline practice and separated between-institution from teacher-relative associations.

Two of five primary tests met the Holm-adjusted criterion. The stress-related product of between-institution coefficients was negative (-0.099; pointwise 95% confidence interval [-0.184, -0.033]). The positive digitalization-stress association was weaker among teachers with higher digital self-efficacy relative to colleagues (interaction coefficient -0.101; [-0.168, -0.037]) but remained positive at the low, average, and high self-efficacy levels examined. The adjusted digitalization-practice association, appraisal-related coefficient product, and appraisal interaction did not meet the corrected criterion.

The findings connect the organization and experience of work with reported AI practice, emphasizing perceived digital work pressure and relative self-efficacy while complementing accounts centered on technology acceptance.