以Moodle日誌探索英語授課修課生之數位學習歷程

An Exploration to Students' E-learning Portfolio upon Using Moodle in the Context of English-Medium Instruction

英語授課為我國高教重點發展項目之一，許多課程以Moodle為教學平台，本研究依據Moodle使用者日誌（log），以資料驅動的概念進行統計與視覺化分析，探索英語授課修課生的數位學習歷程，並以同課程的中文班為對照組。研究結果顯示兩班的Moodle使用態樣相近，學生平均每週登入1.5～1.6次，約70％的登入集中於30％的時段，上課或交作業當日與前一日為登入高峰，全學期有4個學習循環周期。Moodle登入次數與課堂缺席次數呈現中度至高度負相關，與考試成績則為中度正相關。期中考成績、缺席次數、Moodle登入次數、同儕等變數可用於學期成績是否及格的羅吉斯迴歸建模，作為期中預警之用。本研究以巨觀角度分析班級的學習態樣，建議於未來進行個別使用者的微觀研究，而各校Moodle管理者可建置學習指標，輸出相關的學習績效值供教師參考。

English-medium instruction (EMI) has been a focal point of higher education in Taiwan. Many EMI courses use Moodle as the teaching platform. Through statistical analysis and data visualization, this data-driven study utilized the Moodle log to explore students' e-learning portfolios of a two-class course, one in EMI and the other in Chinese-medium instruction. The findings show that the two classes had similar patterns of using Moodle. Each student logged into Moodle for 1.5-1.6 times weekly on average. Around 70% of the logins occurred in 30% of the time slots. The activities peaked either on the day or on one day ahead of the class hours and assignment deadlines. The log file revealed four learning cycles throughout the semester. The number of Moodle logins negatively correlates with that of classroom absences at a medium to high level, but positively correlates with the exam grades at a medium level. Variables such as midterm exam grades, number of classroom absences, number of Moodle logins, and binary peer support are significant in the pass-fail logistic regression model for early warning. This study focused on the macroscopic learning pattern, by class. Future research can focus on individual learners' microscopic analysis, and that Moodle administrators develop learning performance indicators for instructors' reference.