Adaptation of Learning Paths in Online Programming: Clustering and Descriptive Analysis in Engineering

The digital age has transformed education, driving the need to adapt teaching in virtual environments. In online undergraduate courses, we recognize that the automatic adaptation and flexibility of learning paths are crucial to addressing the diverse needs of students and their training. Automation in the sequence of contents, tasks, and assessments is presented as a solution to improve the educational experience in engineering careers. In this context, the instructional designs of online courses require mechanisms that automatically adjust to each student's needs to achieve the learning outcomes proposed for a given course. Therefore, creating student profiles based on their interaction with learning management platforms (LMS) and academic performance is crucial in developing adaptive learning paths. This is especially relevant in programming courses, where students face unique challenges, such as the need for significant autonomous work that necessitates using adaptive learning paths. With the primary objective of designing an adaptive programming course for engineering students that incorporates multiple learning paths based on profile characterization, this paper aims to address the following question: What are the student profiles in a programming course for an online engineering career? To this end, unsupervised learning techniques, such as clustering, were employed to categorize students based on patterns of LMS use behavior and academic performance associated with an existing instructional design for an online Introduction to Programming course. Dimensionality reduction techniques and descriptive analysis of the clusters were also utilized to gain a deeper understanding of the analysis variables and subsequently identify learning paths. The anticipated results could improve the construction of adaptive and flexible learning paths. Implementing these findings has the potential to significantly enhance student retention and academic performance in online programming courses, thereby making a tangible impact on online education and engineering.