Implementing the LSEESC Methodology with AI-Guided Support for Active Learning in Engineering Physics Courses

In this research-to-practice full work, we will explore how innovative educational strategies can improve students' learning experiences. Specifically, we will focus on the early development of skills in engineering students, such as numerical manipulation for problem-solving, proficient spreadsheet management, experimental implementation, and mastery of mathematical modeling complemented by informed decision-making with graphics understanding and interpretation. The effective utilization of available technological resources helps to enhance the student learning experience and prepare them for the challenges in their engineering professional practice. Physics courses often present high rates of low academic performance, but these innovative approaches can enhance active student engagement and help nurture essential skills. However, integrating technology into innovative teaching methods can have implications; one seen recently is the use of Generative Artificial Intelligence (G-AI) in many instruction courses; as a technology of easy access and manipulation, it can result in a confused tool if students don't have the appropriate instruction in its use. Therefore, educating students on the ethical and proper use of AI technology is crucial, guiding them in formulating relevant questions and accurately interpreting the results. To achieve this, we used the Learning Science and Engineering with Electronic Spreadsheets Cycle (LSEESC) methodology, which has been rigorously tested across various contexts over the last four years and has shown promising results. We will support its implementation with comprehensive guidance and use Artificial Intelligence as a prompt/chat tool (e.g., Chat GPT) as a supplementary tool. This strategy aimed to help students develop effective study habits for independent study sessions and promote autonomous learning. We examine how students ask for help from AI, including their choice of words and language. We will also investigate the effects of using casual or informal language and suggest guidelines for constructing practical questions. Our goal was to assist students in obtaining a better understanding and explanations of physics concepts and problem-solving exercises with the help of AI combined with a well-defined methodology, like the LSEESC, which promotes the development of skills like numerical manipulation, mathematical modeling, variables understanding, graphics generation, interpretation, and understanding. The learning intervention considers N = 140 students for working on the LSEESC methodology and using G-AI in physics courses during spring 2024, with students in the first year of academic engineering programs. The topics to evidence the learning process are kinematics and electrics. The results were analyzed with the Normalized Conceptual Gain (NCG) given a value of = 0.75 and 0.67, the Concentration Factor (CF), Rasch's index, and qualitative students' question formulation. This is particularly important because, in some instances, the results provided by G-AI were incorrect owing to the system's inability to detect specific implications that a human can easily recognize; the teacher explained the confused results presented by G-AI and student received the corresponding feedback about these issues with artificial intelligence.