Citra Kurniawan, Punaji Setyosari, Waras Kamdi, Saida Ulfa
In this research, electrical engineering students’ visual-verbal preferences were modelled using the k-means clustering method. The data collected included the level of initial ability of students, student demographic information and student learning preferences. Data were processed using the k-means clustering method, which divides data into several groups or clusters based on the similarity of data attributes. The study identified five clusters, viz. 1) Cluster 0 - informatics, intermediate, male, no, verbal; 2) Cluster 1 - broadcasting, intermediate, male, no, verbal; 3) Cluster 2 - informatics, master, male, no, visual; 4) Cluster 3 - informatics, teachers, men, yes, visual; 5) Cluster 4 - broadcasting, intermediate, male, no, visual. The k-means clustering method is iterative and required six iterations to converge onto a stable solution. Findings indicate that k-means analysis can be used to model student data. Student modelling is essential for learning strategies that when appropriate to the student model can help students to get better outcomes. © WIETE 2018.
Sekolah Tinggi Teknik Malang, Malang, East Java, Indonesia; State University of Malang, Malang, East Java, Indonesia; Department of Education and Technology, State University of Malang, Malang, East Java, Indonesia; Department of Mechanical Engineering Education, State University of Malang, Malang, East Java, Indonesia; Department of Educational Technology, State University of Malang, Malang, East Java, Indonesia
