Language teachers’ AI literacy: A psychometric study based on the ED-AI framework

Texto completo

Artificial Intelligence (AI) is reshaping language education, making AI literacy crucial for teachers to engage critically and effectively with this technology. Nonetheless, most existing assessments target students or general users, leaving a gap in measuring AI literacy within language teacher education. This study sought to develop and validate the Teachers’ AI Literacy Scale (TAILS), grounded in the ED-AI literacy framework, which comprises six dimensions: knowledge, evaluation, collaboration, contextualization, autonomy, and ethics. The scale was tested with preservice English language teachers through two phases: exploratory factor analysis (EFA) with 165 participants and confirmatory factor analysis (CFA) with a separate sample of 227. Results confirmed a six-factor structure with high internal consistency (Cronbach’s α values > 0.90) and acceptable model fit indices (Chi-square/df = 1.766, RMSEA = 0.058, SRMR = 0.054, TLI = 0.908, CFI = 0.919), demonstrating strong validity and reliability. Each dimension aligned clearly with the competencies required for AI-integrated language teaching. The TAILS is a psychometrically robust, context-specific instrument for assessing AI literacy in language teacher education. This study bridges the gap between theoretical frameworks and practical assessment, offering a foundation for curriculum development, professional training, and policymaking. Its application supports the preparation of AI-competent educators equipped to navigate the ethical, pedagogical, and technological demands of the digital classroom.

Highlights:

• We developed the Teachers’ AI Literacy Scale (TAILS) grounded in the ED-AI literacy framework.
• The TAILS consists of six dimensions: knowledge, evaluation, collaboration, contextualization, autonomy, and ethics.
• The TAILS is to assess language teachers’ AI literacy.
• Exploratory and confirmatory factor analysis confirmed strong construct validity.
• The scale demonstrated high reliability with Cronbach’s α values > 0.90.