A Comparative Study of AI-Driven and Human Tutoring Models for Coding Instruction in Initial Teacher Education

The study investigates the impact of human tutor support versus Generative AI (GenAI) support on preservice teachers’ learning of basic programming concepts, explicitly situating GenAI in the context of preservice teacher education rather than conventional computer science cohorts. The study follows a quasi-experimental research design. 70 preservice teachers from a teacher education program at a UK university were divided into two groups: a con-trol group that received traditional in-class support from a human tutor, and an experimental group that received equivalent support from a GenAI system during structured sessions. Quantitative data from the online survey was analyzed using descriptive and inferential statistics. Cognitive engagement, enjoyment and expected performance were found higher for the control group while satisfaction was found higher for the experimental group. No significant differences were found for self-efficacy, competence support, behavioral engagement, emotional engagement, and intention to use. The findings suggest that human tutors foster deeper cognitive engagement and enjoyment, while GenAI enhances satisfaction and autonomy. This complementarity suggests that integrating both forms of support can improve programming education in teacher education programmes by leveraging the unique affordances of each.