An Exploratory Study on the Impact of a Self-Explanation-Based Adaptive Chatbot on Learning Strategies

Abstract

This study developed a programming‑learning support system that blends the self‑explanation strategy with generative artificial intelligence (Gen AI). The system automatically generates adaptive questions from students’ submitted code and provides instant feedback through a chatbot to guide self‑explanation and reflection. A 16‑week experiment conducted in a Python course showed that students with high chatbot interaction significantly outperformed their low‑interaction peers in learning‑strategy gains, especially in the use of rehearsal strategies. Correlational and regression results further confirmed a positive influence of interaction frequency on strategy improvement. The findings show that Gen AI assistance can enhance programming learning and promote strategic learning behaviors, highlighting its practical value for adaptive instruction.