Is there a difference in cognition and behavior among learners with different learning styles when using GenAI-assisted learning? An empirical analysis based on programming learning

With the advancement of generative AI (GenAI), its integration into education is gaining attention. Investigating the cognitive and behavioral differences of learners with different learning styles during the collaborative programming process with GenAI can promote the intelligent transformation of educational models and personalized teaching. This study focuses on programming learning, establishes a programming learning model supported by GenAI, and conducts experiments based on this model to test the effectiveness of GenAI-assisted programming education and analyze the cognitive and behavioral differences of learners with different learning styles when using GenAI. The research found: (1) Collaborative programming with GenAI improves learners' academic performance and programming efficiency; (2) The cognitive network differences of learners with different learning styles in human-computer collaborative programming are insignificant; (3) Overall behavior sequence analysis shows that learners exhibit a typical behavioral pattern centered on “asking-verifying-correcting” during the collaborative programming process with GenAI; (4) There are differences in the behavior sequences of learners with different learning styles in human-computer collaborative programming. Therefore, teachers understand the behavioral differences to design precise personalized plans and foster effective questioning strategies for better use of GenAI tools.