Can Generative Artificial Intelligence Outperform Self-Instructional Learning in Computer Programming?: Impact on Motivation and Knowledge Acquisition

Generative artificial intelligence tools, such as Microsoft Copilot, are transforming the teaching of programming by providing real-time feedback and personalized assistance; however, their impact on learning, motivation, and cognitive absorption remains underexplored, particularly in university settings. This study evaluates the effectiveness of Microsoft Copilot compared to instructional videos in teaching web programming in PHP, implementing a quasi-experimental design with 71 industrial engineering students in Chile, divided into two groups: one using Microsoft Copilot and the other following instructional videos, with pre- and post-tests applied to measure knowledge acquisition while surveys based on the Hedonic-Motivation System Adoption Model (HMSAM) assessed cognitive absorption (enjoyment, control, immersion, curiosity) and technology acceptance (perceived usefulness, ease of use, and intention to adopt). The results show that, while both methods improved learning, students who used instructional videos achieved greater knowledge gains, higher levels of curiosity, and a stronger intention to continue using the technique, suggesting that instructional videos, by providing structured explanations and reducing cognitive load, may be more effective in the early stages of programming learning. In contrast, AI tools could be more beneficial in advanced stages where students require adaptive feedback, providing empirical evidence on the comparative effectiveness of AI-based and video-based instruction in teaching programming and highlighting the importance of balancing structured learning with AI-driven interactivity, with the recommendation that educators integrate both approaches to optimize the learning experience, using videos for initial instruction and AI tools for personalized support.