(Special Issue) Improving Retention Rate Through Educational Data Mining: The Design of Placement Program for Newly Enrolled Students Author: Yung-Hsiang Hu     (Center for General Education, National Yunlin University of Science and Technology), Hui-Yun Yu (Department of Business Administration, National Yunlin University of Science and Technology)

Vol.&No.：Vol. 65, No. 4
Date：December 2020
Pages：31-63
DOI：10.6209/JORIES.202012_65(4).0002

Abstract：
Retention rate is a key indicator of university governance. However, identifying key courses that influence first-year students’ termination of learning and improve their performance during the first semester is critical. In recent years, offering pathway courses during the summer semester has become a common practice for universities. Therefore, this institutional research employed educational data mining analysis and pathway courses to improve retention rate and student success. The data analyses comprised classification and regression trees, the Wilcoxon rank-sum test, k-means clustering, and descriptive statistics. This study first analyzed 22,750 educational big data points from 2,135 freshmen in the sample college from the academic year of 2015 to 2017. Subsequently, decision tree analysis was employed to identify key courses that predicted student suspension, dropping out, or transfer. Thereafter, two pathway courses and remedial teaching were offered to freshmen in the summer to learn online. Finally, this study tracked the success of freshmen and evaluated the effects of the two pathway courses. The major findings suggested the following: (1) Physics (I) and Calculus (I) are key courses, and students who failed both courses were 5.5 times more likely to suspend their studies, drop out, or transfer than was the total student population; (2) The pass rate of formal courses for students who had not watched the audiovisual course was between 50% and 63%, much lower than the total student population rate of 83% to 94%, and only the Calculus (I) gateway course could improve learning readiness; (3) Online supplementary teaching was found to promote the academic performance of freshmen in Physics (I); however, no significant differences were observed in Calculus (I). Moreover, the study group improved students’ academic performance in Calculus (I); however, no significant differences were observed in Physics (I); (4) Compared with the previous year, the retention rate of the sample college increased by 48.07%. Finally, the researchers proposed suggestions for the gateway course’s follow-up application. To conclude, this study may be of importance in explaining the effectiveness of gateway courses, in addition to providing university authorities with a better understanding of how retention rate can be improved through educational data mining and institutional research.

Keywords：enrollment management, educational data mining, retention rate, pathway courses

《Full Text》