Teachers' competence, self-efficacy, and their attitudes toward generative AI in education: a correlational study

The rapid integration of Generative Artificial Intelligence (GenAI) offers new opportunities and challenges for education, with teacher adoption being a key factor for successful implementation. However, the psychological factors influencing teachers' attitudes toward these tools remain poorly understood. This study explores the relationship between K-12 teachers' digital competence, their self-efficacy for AI integration, and their attitudes toward using GenAI in education.

A quantitative, cross-sectional survey design was used. Data were collected from 352 K-12 teachers in a large urban school district in Indonesia through an online questionnaire. The survey assessed four constructs: General Digital Competence (GDC), AI-Specific Competence (AISC), Teacher Self-Efficacy for AI Integration (TSE-AI), and Attitudes Toward Generative AI in Education (ATGAI-E). Data analysis involved Pearson's correlations and hierarchical multiple regression.

The analysis showed strong positive correlations among all variables. Attitudes toward GenAI had the strongest correlation with TSE-AI (r = 0.75, p < 0.001), followed by AISC (r = 0.69, p < 0.001) and GDC (r = 0.58, p < 0.001). The hierarchical regression model explained 62% of the variance in teachers' attitudes (Adjusted R ² = 0.61). After accounting for demographic factors, TSE-AI was the most significant predictor (β = 0.48, p < 0.001), with AISC (β = 0.25, p < 0.001) and GDC (β = 0.11, p = 0.015) also contributing.

The findings indicate that while both general and AI-specific skills are important, a teacher's self-efficacy, their confidence in their ability to use GenAI effectively, is the most powerful predictor of their attitude. This suggests that successful GenAI adoption depends more on psychological empowerment than on technical training alone. Professional development programs should therefore focus on building teachers' confidence through hands-on practice, peer modeling, and supportive feedback to encourage positive engagement with the GenAI classroom.