Beyond school support: Internalizing the school innovation support system in teachers’ digital teaching innovation

Promoting teachers’ digital teaching innovation has become a key challenge in AI-enhanced junior secondary education, particularly under high-stakes examination pressures. This study uses structural equation modeling, Shapley value decomposition, and fuzzy-set qualitative comparative analysis to examine how the school innovation support system shapes teachers’ digital teaching innovation. The findings show that the school innovation support system does not directly translate into digital teaching innovation. Instead, its influence is mainly transmitted through teachers’ AI readiness and teaching self-efficacy, with AI readiness serving as the strongest pathway. Shapley value decomposition shows that the school innovation support system is not a purely top-down mechanism. Vertical leadership support plays a slightly stronger enabling role, while horizontal peer support remains almost equally consequential in sustaining teachers’ digital teaching innovation. The configurational results further show that high digital teaching innovation can emerge through multiple pathways, while AI readiness and teaching self-efficacy consistently serve as core conditions.