Leveraging generative AI to enhance sixth graders' scientific argumentation quality on the greenhouse effect topic

This case study aimed to explore how a learning module designed based on the Argument-Driven Inquiry (ADI) instructional model with generative AI integration facilitated sixth-grade students' written scientific argument quality in the context of the greenhouse effect, their interaction with the chatbot, and their perceptions of this experience. Conducted after school hours at a public middle school in Mersin, Turkey, the study involved five high-achieving sixth-grade students over three weekly 120-minute sessions. During each session, participants investigated different aspects of the greenhouse effect results, engaging in various simulations, text materials, and chatbot interactions to form their scientific arguments. Data, including participants' written scientific arguments, chatbot conversation logs, and perception forms, were analyzed using content analysis to assess argument quality and participant experiences. Findings showed that the chatbot's feedback and guidance improved argument quality, particularly in refining claims, evidence, and reasoning. Although participants initially struggled with interpreting complex responses, the chatbot generally produced contextually appropriate responses. Participants reported enhanced topic understanding and new knowledge acquisition. Due to technological limitations at the school, the small sample size led the study to focus on individual written argumentation, offering deep insights into chatbot efficacy while limiting the transferability of findings to larger populations.