Predictive Learning Models Based on Student Engagement on Engineering Education

In the realm of education, the prevailing focus on rote memorization often eclipses the necessity for a holistic and multifaceted approach to student learning. While factual recall remains crucial, more than an overreliance on this method is needed for the enduring comprehension and practical application of information, particularly in the context of engineering education. This study delves into the Mechanical Engineering Practice & Graphics course (ME1403) at UTSA, which embraced a teaching model aligned with Bloom's taxonomy. This framework categorizes cognitive skills into ascending levels of complexity, advocating for a comprehensive strategy that nurtures critical thinking, problem-solving, and practical knowledge application. During the Fall 2023 semester, the ME1403 course focused on instructing students to utilize SolidWorks CAD software—a tool integral to electronic design. As the semester progressed, a deliberate transition towards a more student-driven approach unfolded, challenging individuals to craft original designs autonomously with diminishing guidance. While acknowledging areas for refinement in student engagement, incorporating Bloom's Taxonomy-inspired framework in the Mechanical Engineering Practice & Graphics course at UTSA signals a pivotal shift towards a more comprehensive and effective educational paradigm. Additionally, our analysis explores the correlation between grades of homework, quizzes, exams, and the grades of the final project involving 120 students across four sections. These grades formed the data points for a predictive learning model to anticipate future educational needs and guide course development, thereby contributing to a more adaptive and responsive educational environment. The study found the learning model to hold great promise for predicting the success of students' final project grade. However, given the relatively small number of datapoints used for the study, the model's effectiveness cannot be completely validated.