Predicting Licensure Examination Performance using Decision Tree

Abstract

Higher Education Institutions consider the performance of their students in licensure examination as a measure of high-quality of education they offer. One factor that contributes to the performance of the students for their licensure examination is their academic performance in the HEI. The paper is about predicting the performance in the licensure examination of engineering students. Data collection, data cleaning, data selection, and data transformation are all part of the pre-processing stage. During forecasting step, ID3 decision tree algorithm is applied to generate the model in predicting the result of the licensure examination. WEKA tool was used to generate the ID3 decision tree. For this research, 10 folds cross-validation for model assessment was used.   Based on the results, ID3 algorithm correctly classified 59 instances of the data set: 22 were correctly classified as True Negative and 37 were correctly classified as True Positive. However, there were 16 incorrectly classified instances: 9 were classified as False Positive and 7 were classified as False Negative. The prediction model also shows that an outstanding performance in the Professional courses can help the students successfully pass the licensure examination. The ID3 decision tree algorithm has an accuracy performance of 81.94%. The research may help increase the chances of students passing their licensure examination performance by focusing on professional courses. The identification of what technical courses contribute most to the success of the student in passing the licensure examination. This research may also help higher educational institution in creating effective intervention strategies to improve the success rate of students in the licensure examination.