（專題）以教育大數據分析驅動入學管理機制開設新生銜接課程提升就學穩定度之研究 作者:國立雲林科技大學通識教育中心胡詠翔、國立雲林科技大學企業管理系俞慧芸

卷期：65卷第4期
日期：2020年12月
頁碼：31-63
DOI：10.6209/JORIES.202012_65(4).0002

摘要：
不該只看註冊率，在少子女化衝擊下，維持學生入學後的就學穩定度是學校永續經營須重視的關鍵；尤其，新生必須面對與《十二年國民基本教育課程綱要》和高中學習經驗截然不同的院系本位課程與教材內容，大學端該如何提高大一學習經驗，找出影響新生改變學習旅程規劃的關鍵學科，透過入學管理機制開設銜接課程，以維持就學穩定度，成為未來高等教育的重要命題。近年，有大學嘗試自辦暑期銜接課程以解決問題。本研究首先分析個案學校104至106學年某學院2,135位新生，計22,750筆教育大數據資料，透過決策樹分析找出影響新鮮人休學、退學或轉系之關鍵的大一上學期課程，再辦理暑期小規模非公開遠距課程（含補救教學與讀書會），追蹤效果進行機制評估。結果發現：一、物理（I）與微積分（I）這兩門院必修是關鍵課程，且兩門皆不及格學生的休學、退學或轉系機率是原母體的5.5倍；二、未觀看銜接課程教材者開學後的關鍵課程及格率介於50%～63%間，遠低於其他落在83%～94%，且又以微積分的銜接課程具統計上顯著提升學習準備度效果；三、線上補救教學有助於物理（I）的學業表現，讀書會則有助於微積分（I）的學業表現；四、個案學院就學穩定度較前一年提升48.07%。最後，本研究提出具體建議供後續研究與規劃課程參考。

關鍵詞：入學管理機制、教育大數據分析、就學穩定度、銜接課程

《詳全文》

(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