Guided Learning with AI: A Didactic Strategy Using Sequential Tutoring via ChatGPT for Object-Oriented Programming

This study investigated the pedagogical effectiveness of generative artificial intelligence (AI) tools—specifically ChatGPT—in the teaching of object-oriented programming (OOP) through a structured, AI-assisted sequential tutoring strategy. The research adopted a mixed-methods experimental design to compare learning outcomes between two student groups: one that followed traditional instructional methods and another that received AI-mediated guidance through a didactic model based on progressive scaffolding. The methodology included pre- and post-tests, perception surveys, and performance evaluations across theoretical and practical tasks. Results showed that the AI-assisted approach enhanced students’ practical programming skills, engagement and perceived self-efficacy. However, theoretical content understanding remained comparable between groups, with a slight advantage observed in the control group. The study also synthesised findings from 33 peer-reviewed sources, framing the intervention within established pedagogical theories such as Bloom’s Taxonomy, constructivism, and adaptive learning. The integration of ChatGPT enabled personalised feedback, real-time error correction and metacognitive reinforcement, particularly during complex coding tasks. Findings supported the adoption of generative AI as a complementary instructional tool rather than a substitute for human teaching. Embedding AI into coherent, goal-driven didactic frameworks allowed for scalable, adaptive and engaging learning environments in computer science education. The study concluded by offering guidelines for future implementations and emphasising the need for continued research on responsible AI integration in programming instruction.