Integração da inteligência artificial no ensino de física térmica: articulando ABP e atividades experimentais no desenvolvimento de uma sequência didática para o Ensino Superior

• Ano de defesa: 2025
• Autor(a) principal: Becker, Daniel de Moraes
• Tipo de documento: Dissertação
• Tipo de acesso: Acesso aberto
• Idioma: por
• Instituição de defesa: Universidade Federal de Santa Maria; Brasil; UFSM; Programa de Pós-Graduação em Educação Matemática e Ensino de Física; Centro de Ciências Naturais e Exatas
• Palavras-chave em Português: Inteligência artificial; TDIC; Ensino de Física; Sequência didática; Ensino Superior; Artificial intelligence; Physics teaching; Didactic sequence; Higher Education; DICT; CNPQ::CIENCIAS HUMANAS::EDUCACAO
• Link de acesso: http://repositorio.ufsm.br/handle/1/36864

Artificial Intelligence (AI), especially generative AI, has rapidly expanded in the media and in students' daily lives since the launch of ChatGPT at the end of 2022. The capability of this technology to create coherent texts in natural language and to generate original images has sparked debates about its impact on society and the ethical boundaries of its application. In the educational context, AI has been integrated at an increasing pace, raising numerous questions for teachers and students about how to use it in ways that maximize its benefits and minimize potential negative effects. In this context, the present work aims to investigate how the articulation between the principles of Project-Based Learning, experimental activities, and an AI-powered chatbot can promote continuous inquiry, active participation, and reflection by students on the content, during the development and implementation of a didactic sequence in Thermal Physics. To this end, Educational Action Research (EAR) was used as the research approach, which organizes planning, action, observation, and reflection in cyclical moments, based on the researcher's proximity to the research context, with a view, among other things, to improving teaching practices. Initially, the main theoretical frameworks underpinning the study were explored, including a discussion about Digital Information and Communication Technologies (DICT) in Physics and Science teaching, the role of AI as a technology, and the assumptions that guide the construction of the didactic sequence. To contextualize the application of AI in Physics and Science teaching, a Systematic Literature Review was carried out to identify relevant works integrating AI published in major journals, academic events, dissertations, and theses. Based on the results of this analysis, a didactic sequence integrating AI was developed, discussing various possibilities, challenges, and the methodological paths taken in its conception. The didactic sequence was implemented in the “Introduction to Physics Laboratory” course of the Physics teaching degree at the Federal University of Santa Maria. Data analysis was conducted in five stages proposed by Yin (2010): compilation, decomposition, recomposition, interpretation, and conclusion. The analysis of this implementation showed that the didactic sequence fostered active student participation in building the proposed project. Although AI assisted in generating initial ideas, practical challenges required dialogue and collaboration among groups, encouraging the collective construction of practical and theoretical knowledge. The combined use of these methodologies and resources promoted continuous and reflexive inquiry into physical concepts, as well as the reliability of the information generated by AI.