Adaptation of the artificial ıntelligence literacy scale into Turkish: A cross-sectional application among healthcare workers, students, and children

Abstract In this study, the Meta-Artificial Intelligence Literacy Scale developed by Carolus et al. (Computers in Human Behavior: Artificial Humans, 1(2), 100014) was adapted to Turkish, and its validity and reliability were tested. This cross-sectional research was conducted with three distinct study groups to assess the scale's applicability and consistency in diverse contexts. Validity and reliability analyses were performed with 1167 participants between the ages of 12–74, representing a broad demographic range. Additionally, the scale’s applicability was further evaluated with 583 children aged 12–18 and 122 healthcare workers. The Meta-AI Literacy Scale comprises four main dimensions—artificial intelligence literacy, artificial intelligence creation, artificial intelligence self-efficacy, and artificial intelligence self-competence—and eight sub-dimensions that cover a wide range of AI-related competencies. Cronbach Alpha and test–retest analyzes of the scale showed high consistency. Test–retest analyses confirmed the stability of the instrument over time. The findings revealed significant variations in artificial intelligence literacy based on age and education level, highlighting the influence of demographic factors. However, no significant differences were observed based on gender. This study concludes that the Meta-Artificial Intelligence Literacy Scale is a valid, reliable, and versatile tool for assessing AI literacy among Turkish-speaking individuals in various settings, including educational and healthcare contexts.