Evaluating Experiential Learning and AI-Supported Scaffolding in Differential Equations Courses for Engineering Students

Abstract

This study presents the implementation and evaluation of an educational strategy based on experiential learning and the regulated use of artificial intelligence tools as support for mathematical reasoning in a Differential Equations course for engineering students. The study was conducted using a comparative quasi-experimental design, in which three groups taught by the same instructor were analyzed: one control group following a traditional instructional approach and two experimental groups that incorporated contextualized activities in real engineering scenarios and the controlled use of artificial intelligence as cognitive scaffolding. The assessment of academic performance and disciplinary competencies was designed in alignment with the learning objectives and institutional competency frameworks of the course, using instruments integrated into the regular evaluation process. The results show consistent differences in favor of the experimental groups in both final grades and criteria associated with complex problem solving, mathematical modeling, and solution interpretation. The analysis suggests that the integration of experiential learning and the regulated use of artificial intelligence promotes a more applied understanding of mathematical concepts, while maintaining students as active agents in the learning process. Overall, the study provides empirical evidence on the feasibility of pedagogical approaches aligned with Education 6.0 in foundational mathematics courses for engineering within Latin American contexts.