Uso do chat GPT no ensino médio aplicado ao ensino de Física

Abstract

This study investigates the effects of using ChatGPT, an artificial intelligence-based tool, as pedagogical support in high school Physics education. The study is grounded in the active methodology Peer Instruction, developed by Eric Mazur, and focuses primarily on teaching thermodynamics concepts, especially those related to heat, temperature, and thermal sensation—topics frequently associated with students’ misconceptions. The research was conducted at the Escola Estadual de Educação Profissional Maria Altair Américo Saboia, in Independência, Ceará, where one class used only the Peer Instruction method and was compared with two others that also employed the same method, but with ChatGPT as a support tool. Conceptual tests were administered before and after the interventions, allowing for the assessment of learning progression and the effectiveness of the approaches. One of the main objectives of this study is to highlight the need to address misconceptions about heat and temperature, which are deeply rooted in everyday language, with historical and cultural origins dating back to the Roman Empire, and are also present in students’ prior knowledge. By combining conceptual explanations with the use of artificial intelligence, the study seeks to promote more meaningful and consistent learning, going beyond rote memorization and valuing scientific understanding and digital literacy. The results suggest that the integration of Peer Instruction with the guided use of ChatGPT can contribute to improved conceptual performance among students and represents a promising strategy to tackle long-standing challenges in Physics education—though it does not replace the importance of student protagonism.