The relationship between primary school teachers' self-efficacy in differentiated instruction, digital fluency and perceptions of artificial intelligence in education

Objective: Curricula are designed within a holistic framework that addresses all dimensions of individual development, taking students' interests, needs, and individual differences into account. In this context, differentiated instruction has become a central approach in contemporary educational practices, and its effectiveness is closely related to teachers' pedagogical and technological competencies. Based on this premise, this study aims to examine primary school teachers' self-efficacy in differentiated instruction in relation to their levels of digital fluency and their perceptions of using artificial intelligence (AI) in education. Materials and Methods: A mixed-methods approach was employed, adopting a sequential explanatory design. The study population consists of classroom teachers working in public primary schools located in three central districts of a metropolitan city in the Eastern Anatolia Region during the 2024-2025 academic year. The quantitative sample included 421 teachers from 35 schools selected through stratified cluster sampling, while the qualitative phase involved interviews with 16 teachers selected via maximum variation sampling. Data were collected using the "Differentiated Instruction Self-Efficacy Scale," "Digital Fluency Scale," "Artificial Intelligence in Education Scale," and a semi-structured interview form, and analyzed using appropriate techniques. Results: Findings indicated that teachers' self-efficacy in differentiated instruction and digital fluency levels were moderate, whereas their perceptions of AI use in education were high. Significant differences were observed across various demographic variables. Furthermore, differentiated instruction self-efficacy was significantly related to digital fluency and AI use perceptions, with these variables explaining 54% of the total variance. Qualitative findings supported and deepened the quantitative results. Conclusion: The findings indicate that there are significant and positive relationships among primary school teachers' differentiated instruction self-efficacy, digital fluency, and perceptions of artificial intelligence use in education. Furthermore, digital fluency and AI-use perceptions were found to be significant predictors of differentiated instruction self-efficacy, jointly explaining a substantial proportion of its variance.