In 2015, the Institutional Research (IR) Project was launched, using big data to track and analyze various accumulated student data to improve Taiwan higher education quality. By using educational data mining (EDM) methods, the future possibility of individual students in facing suspension or dropout crisis can be analyzed, thereby enhancing higher education quality. This study collected years of case university student data, adopted EDM to analyze cases of student suspension and dropout, further achieved in forecasting the future possibility of student suspension or dropout. The study results serve as references for developing an early warning system (EWS). In this study, chi-squared automatic interaction detection (CHAID) was adopted as a decision tree-growing method. The model exhibited a correct classification rate of 84.75%. Case university have 5,718 enrolled students. EDM predictive results revealed that 103 students are likely to drop out. The major reason for dropout was personal and academic factors. This study methods and results can serve as references for universities to develop EWS.
No takes yet. Share an insight, caveat, or question.
Sung-Po Lin (2015) studied this question.