A Study On Factors And Using Machine Learning For Learning Analytics In Computer Programming Of Junior High School Students

Abstract

This research aims 1) to study factors that affect the learning of Computer Programming of Junior High-school students, and 2) to study the Machine Learning Techniques in analytics of factors that affect the learning of Computer Programming of Junior High-school students. Data were used came from the survey of 411 Junior High-school students. The preliminary analysis using the document research and collect from 2 round of experts confirmed and leading the selected factors to Exploratory Factor Analysis (EFA) and then to analytics by 5 Machine Learning Techniques which were K- Nearest NeighborsTechnique, Logistic Regression Technique, Decision Tree Technique, Support Vector Machine Technique, and Naïve Bayes Technique.
Results found that,
1) The Exploratory Factor Analysis (EFA) by Correlation Coefficient found that there were 11 factors of Instructor, 4 factors of learning environment, 5 factors of learning media, and 11 factors of learner which affect to learning Computer Programming.
2) The results of using Machine Learning Techniques found that there were Logistic Regression Technique, Decision Tree Technique, and Naïve Bayes Technique showed the highest accuray in Gender class, and there were Decision Tree Technique, and Naïve Bayes Technique showed the highest accuray in Level of study class.