Abstract
The interplay of factors, including both cognitive and non-cognitive, plays a significant role in the learning patterns of students. However, the majority of the research conducted on such issues mainly puts forward the role of cognitive skills but forgets that a very important role is played by the non-cognitive factor, specifically motivation and emotional intelligence. Therefore, this study focuses on bridging that gap by investigating the combined influence of cognitive and non-cognitive factors on the learning capacities of engineering students during their transition to higher education. A two-year longitudinal study on engineering students of AITAM, Tekele, India was considered in relation to their academic performance, learning preference, and socio-emotional aspects. The approach adopted makes use of predictive analytics. It is deployed here as machine learning algorithms in the form of Logistic Regression (LR), Naive Bayes, k-Nearest Neighbors (k-NN), Decision Trees (DT), and Support Vector Machines (SVM) to classify the learners into very fast, fast, average, and slow learners. The algorithm of k-NN also achieved the highest accuracy classification and showed good robustness for learning the students’ learning rates. This study underscores the combination of new teaching approaches as well as personalized self-learning methods to enhance learning performance, especially for slow learners. Indeed, the outcome gives avenues for much more extensive studies done on large datasets using advanced algorithms which can be applied across a range of educational fields to support tailored learning interventions.
Keywords: Classification, Cognitive Learning, Education, Machine Learning, Non-Cognitive Learning, Student Performance.