Teachers’ digital language literacy in the AI era: framework construction and validation

This study constructs and empirically validates a novel framework for teachers’ digital language literacy (TDLL) in the artificial intelligence (AI) era—the first of its kind to be rigorously validated. Through systematic analysis of 12 international policy documents and a three-round Delphi consultation with 30 experts, a five-dimensional TDLL framework was established: cognition and critique, tools and collaboration, multimodal resource design, data and personalized instruction, and ethics and safety. The Delphi process, coupled with entropy weight calculations, identified ‘tools and collaboration’ (weight=0.236) and ‘ethics and safety’ (weight=0.211) as the most critical dimensions. Empirical validation via a large-sample survey (N=984) confirmed the framework’s high reliability (Cronbach’s α=0.936) and good construct validity (comparative fit index (CFI)=0.921, root mean square error (RMSEA)=0.048). Descriptive analysis revealed teachers were most competent in tool application (M=4.12) but least competent in data-driven personalized instruction (M=3.78), highlighting a key training gap. The study provides a scientifically grounded assessment tool and actionable guidance for targeted teacher professional development in the AI age.