abstract
- In the contemporary educational landscape, the ability to generate information regarding various facets of a student's learning process has become increasingly feasible. Consequently, the effective interpretation of this data and its translation into meaningful insights about student learning has become a paramount concern. Among the diverse educational innovations that have garnered attention in recent times, the implementation of predictive systems utilizing machine learning in engineering science education has demonstrated the potential to yield positive and measurable impacts on different stages of the teaching and learning process. This research aims to evaluate the implementation of machine learning (ML) in engineering science education and selected STEM areas at the university level while identifying the cognitive levels that can be enhanced with the support of this technology. Through a qualitative meta-synthesis research approach, successful applications of ML have been identified in predicting student attrition, monitoring academic performance, and implementing regularization actions. This study offers a guide for evaluating learning and comprehending the scope of ML in engineering and STEM education, drawing comparisons between the educational theory related to achievement scaling in knowledge construction as defined by learning taxonomies and the ML systems proposed by various authors. © 2023 IEEE.