AI literacy development in middle school: Profiles and demographic predictors from an ABCE perspective

This study examines artificial intelligence (AI) literacy among middle school students through the lens of the ABCE framework, which encompasses Affective, Behavioral, Cognitive, and Ethical dimensions. Using a sample of 1,669 students from Shanghai, China, the research validates the AI Literacy Questionnaire (AILQ) as a robust psychometric tool for assessing these four domains. Latent profile analysis identified four distinct AI literacy profiles, Poor, Moderate, Good, and Excellent, revealing that 72.3% of students achieved high literacy levels (Good or Excellent), while 27.7% remained in lower proficiency groups (Poor or Moderate). Key demographic findings highlight the role of gender and grade level in AI literacy development. Specifically, males had significantly higher odds of being in the “Excellent AI literacy” group compared to the “Poor AI literacy” group. Grade level also significantly predicted profile membership, with Year 10 and Year 11 students showing higher odds of achieving “Excellent AI literacy” compared to the “Poor AI literacy” group. These findings highlight the multi-faceted nature of AI literacy and its varied distribution among students. The validated ABCE framework and AILQ provide valuable tools for holistic assessment and curriculum design. The identified demographic influences underscore the need for differentiated and equity-focused instructional strategies to ensure all students are prepared for an AI-driven world. • This study validates the AI Literacy Questionnaire (AILQ) as a reliable tool to assess middle school students’ AI literacy across affective, behavioral, cognitive, and ethical (ABCE) dimensions. • This study also identifies distinct AI literacy profiles and examines how demographic factors such as gender and academic year predict these profiles. • Results of this study show that policymakers and school leaders should consider equity-focused strategies, particularly gender-responsive and grade-level-specific interventions, to ensure inclusive development of AI literacy skills among all students.