Hai-Hua Wu, Lei Yang
Generative artificial intelligence (GenAI) is increasingly integrated into second language (L2) education, with its multimodal capabilities supporting language learning. For pre-service teachers, developing the literacy to harness GenAI’s potential is essential. However, psychometrically sound instruments to assess such literacy remain scarce, and its implications for teacher well-being, particularly in under-resourced regions, are underexplored. To address these gaps, this study validated a GenAI literacy scale for pre-service L2 teachers and examined its impact on well-being. A 22-item scale was constructed among 480 participants from universities across China, with the well-being analysis focusing on 253 pre-service L2 teachers from Ningxia, one of China’s less developed provinces. The scale comprises four factors: Knowledge, Skills, Traits, and Motives. Using Mplus 8.7, a comparison between exploratory structural equation modeling (ESEM) and confirmatory factor analysis showed that ESEM provided a more accurate psychometric representation. The scale also demonstrated strong validity, reliability, and gender invariance. Subsequent analysis using partial least squares structural equation modeling via Smart PLS 4 revealed that GenAI literacy accounted for 37.1% of the variance in well-being. Notably, only Skills and Motives had a significant positive effect, while Knowledge and Traits showed no impact. These findings contribute to both theory and practice by clarifying the structure and psychological relevance of GenAI literacy in teacher education and by highlighting its role in supporting the well-being of pre-service teachers in under-resourced regions.