使用序列分群法揭示學生學習策略並探討其與認知負荷及學習成效之相關性;Using Sequence Clustering to unveil students’ learning strategies and explore their relationship with Cognitive Load and Learning Performance

NCUIR > College of Electrical Engineering & Computer Science > Graduate Institute of Computer Science and Information Engineering > Electronic Thesis & Dissertation > Item 987654321/83909

Please use this identifier to cite or link to this item: http://ir.lib.ncu.edu.tw/handle/987654321/83909

Title:	使用序列分群法揭示學生學習策略並探討其與認知負荷及學習成效之相關性;Using Sequence Clustering to unveil students’ learning strategies and explore their relationship with Cognitive Load and Learning Performance
Authors:	康景翔;Kang, Ching-Hsiang
Contributors:	資訊工程學系
Keywords:	認知負荷;序列分析;學習策略;學習成效;Cognitive Load;Sequence Clustering;Learning Strategy;Learning Performance
Date:	2020-06-30
Issue Date:	2020-09-02 17:39:16 (UTC+8)
Publisher:	國立中央大學
Abstract:	學生在測量認知負荷程度時必須使用問卷進行測量，而認知負荷問卷實務上需要間隔兩周至三周一次進行施測，頻繁地施測也會造成學生的問卷疲勞，且收集問卷後還需要進行統計分析，以至於導師在得知學生的認知負荷負載程度時往往已近學期尾聲，許多干預或輔導措施已不具即時性，難以再進行拯救輔導等等相關措施。除此之外，一些與時間性有關的題目，可能會因為學生在反思過程中記憶模糊或主觀意識影響導致作答有誤，其答案的正確性若能透過線上系統的日誌分析，則會更加精確且可靠。換句話說，如果能夠從學生操作學習管理系統(Learning Management System, LMS)的learning logs就揭露學生認知負荷(cognitive load)的狀況，對於認知負荷的測量是十分即時且快速，且對於教師授課內容的調整、預防以及者干預措施也可獲得即時性的效益。本研究收集了113位大學生在LMS留下的learning logs，並將learning logs編碼成序列以保持學習歷程的完整性，並將類似的序列分群以檢測學生的學習策略。分析的結果一共得到五種常見的線上學習策略，並找到一種學習策略可以替代測量(proxy measure)學生的認知負荷程度，本研究同時驗證了學生學習策略與學習成效之間的關聯性，結果表明部分學習策略與學習成效具有正向且顯著的關聯。最後我們得知，學生想要獲得較佳的學習成效，如何正確搭配教材影片是不可或缺的一環，研究發現觀看影片以及做線上測驗的順序是非常重要的，若學生跳過影片直接進行線上測驗，表示學生在課堂中已經累積過多的認知負荷，需要進行即時的調整，反之，若學生正確的使用課程的教材搭配影片學習，循序理解教材內容，對學習是有正向助益的。 ;Cognitive load must be measured with a questionnaire, but the questionnaire measurement can only conduct once every two or three weeks in practice, and frequent measuring will also cause students are aware of questionnaire fatigue. After collecting the response of the questionnaire, it takes time to perform a statistical test for understanding students′ real reflection. Therefore, when teachers know the degree of cognitive load of the students, the semester is almost at the end. Teachers have difficulty giving timely interventions for counseling accordingly. Besides, for some questions related to time in the questionnaire, if students answer through reflection, may lead to wrong answers because of the recall process or subjective consciousness. However, it is more reliable and real-time if we could unveil students′ cognitive load by using learning logs generated from the learning management system, and the adjustment of teaching content, interventions and counseling accordingly can also obtain immediate benefits. In this study, we conducted a system programming course with 113 college students and collected students′ learning logs from a learning management system. We encode learning logs into sequences to maintain the integrity of the learning process, and cluster similar learning sequences to detect patterns in students′ learning strategies. Adopting a sequence cluster analysis, we extracted five learning strategies from the collected learning logs. Moreover, experimental results demonstrated that those learning strategies could unveil the degree of cognitive load of students. This study also estimated the correlation relationships between students′ learning performance and students′ learning strategies. The results present two factors that are under a positive and significant relationship. Furthermore, if students want to get better learning performance, how to properly match the course materials and videos is very important. The study found that the order that students watch videos and take quizzes is a critical factor. Students who skip the videos and take quizzes directly with higher cognitive load in the classroom, and it needs to give intervention immediately. On the contrary, if students correctly use the course materials and videos to learn, and understand the teaching content step by step, it will be of positive help to learning.
Appears in Collections:	[Graduate Institute of Computer Science and Information Engineering] Electronic Thesis & Dissertation

Files in This Item:

File	Description	Size	Format
index.html		0Kb	HTML	93	View/Open

社群 sharing

Loading...