Investigating the correlation between candidate teachers’ acceptance of generative artificial intelligence and artificial intelligence literacy across various disciplines

Key findings from this study

• The study found that GenAI acceptance among prospective teachers did not differ by gender or daily internet use but varied significantly by department, grade level, AI tool usage, and self-perceived proficiency.
• The researchers demonstrate that AIL levels showed significant differences across gender, department, grade, tool usage, and proficiency, with higher scores among those who received formal AI training.
• The authors report that qualitative analysis identified nine factors affecting GenAI acceptance and thirteen factors influencing AIL, including ethical considerations, discipline-specific applications, and understanding of machine learning relationships.

Overview

This study examines the relationship between prospective teachers' acceptance of generative artificial intelligence and their artificial intelligence literacy levels across different academic disciplines. The research addresses a gap in understanding how demographic and experiential variables influence these constructs in teacher education contexts. An explanatory sequential mixed methods design combines quantitative data from 723 prospective teachers with qualitative data from 48 participants. The investigation focuses on identifying factors that shape both GenAI acceptance and AIL, with attention to discipline-specific variations. The study responds to limited research on how teacher candidates across various fields engage with AI technologies and develop related competencies. The authors examine whether gender, department, grade level, AI tool usage, and self-perceived proficiency create meaningful differences in acceptance and literacy levels. The research also explores how formal AI training influences these outcomes among future educators.

Methods and approach

The study employed an explanatory sequential mixed methods approach, collecting quantitative data through an Information Form, a GenAI Acceptance Scale, and an AIL Scale administered to 723 prospective teachers. Follow-up qualitative data came from interviews with 48 prospective teachers. Parametric tests including independent samples t-tests and ANOVA examined quantitative differences across demographic and experiential variables. Pearson correlation analysis assessed relationships between GenAI acceptance and AIL scores. The researchers used MAXQDA software to identify thematic factors influencing both constructs from qualitative interview data. Variables examined included gender, academic department, grade level, daily internet use, AI tools used, self-perceived proficiency, and prior AI training. The sequential design allowed qualitative findings to clarify and expand upon quantitative results. Thematic analysis revealed specific factors affecting acceptance and literacy levels beyond what statistical comparisons could demonstrate. The combination of standardized instruments and open-ended interviews provided both breadth and depth in understanding prospective teachers' engagement with AI technologies.

Results

GenAI acceptance levels showed no significant differences by gender or daily internet use but varied significantly by academic department, grade level, AI tools used, and self-perceived proficiency. AIL demonstrated significant differences across gender, department, grade level, tool usage, and proficiency level, with higher scores among prospective teachers who received formal AI training. Qualitative analysis identified nine factors affecting GenAI acceptance: daily use, problem-solving applications, learning applications, mentor usage, assistance from others, proficiency level, productivity gains, discipline-specific skills, and task efficiency. Thirteen factors influenced AIL levels: understanding AI importance, ethical considerations, AI support in daily life, ability to explain AI, understanding relationships between deep learning and machine learning, big data knowledge, AI decision-making processes, knowledge of AI tools, interpretation of AI technologies, critical evaluation skills, data privacy importance, machine learning knowledge, and evaluation of AI applications within specific disciplines. The qualitative data revealed discipline-specific variations in how prospective teachers perceive and engage with GenAI tools. Acceptance and literacy emerged as related but distinct constructs influenced by overlapping yet different factor sets. The findings indicate that experiential variables such as tool usage and proficiency carry greater weight than demographic characteristics in predicting both outcomes.

Implications

The findings suggest that teacher education programs should develop discipline-specific AI literacy training rather than generic approaches, given the significant departmental variations observed. Formal AI training demonstrably increases literacy levels, indicating that structured educational interventions can effectively develop these competencies among prospective teachers. The absence of gender differences in GenAI acceptance contrasts with gender differences in AIL, suggesting acceptance may depend more on practical experience than foundational knowledge. Programs should emphasize hands-on engagement with AI tools and foster self-perceived proficiency to enhance both acceptance and literacy. The factors identified provide actionable targets for curriculum development and professional preparation initiatives. Educators designing teacher training programs can prioritize elements such as problem-solving applications, ethical considerations, and understanding of machine learning relationships. The study underscores the importance of addressing data privacy and critical evaluation skills as core components of AI literacy for future teachers. Given the proliferation of AI in educational contexts, preparing prospective teachers with both technical understanding and practical acceptance becomes essential. The discipline-specific nature of influences suggests that mathematics, science, language, and social science education programs may require tailored approaches to AI integration. These findings contribute to ongoing discussions about equitable and effective AI implementation in education systems.

Scope and limitations

This summary is based on the study abstract and available metadata. It does not include a full analysis of the complete paper, supplementary materials, or underlying datasets unless explicitly stated. Findings should be interpreted in the context of the original publication.