影響線上學習與合作學習學習成就之各項因素探討

Influence of Various Factors on the Learning Achievement under Online Learning and Cooperative Learning

陳建銘(Chien-Ming Chen)

指導教授 ： 鍾斌賢

中原大學/工學院/電子工程研究所/博士(2011年)

https://doi.org/10.6840/cycu201200001

摘要

本研究提出一個線上學習歷程診斷系統，可以藉由資料探勘、集群分析及統計工具，將學生依其線上學習行為、學習成就及思考風格分群，以供教師針對不同學生線上學習行為採取不同輔導機制。實驗顯示適度發送適性化警訊對線上表現優良及中等學生有鼓勵作用，使得學生線上表現更為積極；對於線上表現不佳之實驗組學生，經過持續的適性化警訊發送，除了表現積極之外，其學習成就與不發送警訊之對照組學生比較，結果有明顯進步。進一步考量思考風格與線上學習行為的關係，針對不同思考風格學生，依其線上表線給予適性化警訊，實驗結果亦證實線上表現不佳的實驗組學生，其學習成就比對照組學生有明顯進步。
由於學生於網路學習平台各項學習活動之線上表現有好幾個觀察維度，若純粹觀察這幾個量化之數值，很難了解學生整體的表現如何，若有一能綜合學生各項線上表現，自動計算出一個可以評估學生線上學習表現的分數之演算法，必能大幅度的減輕教師線上督導之負擔。本研究利用主成份分析法，針對學生線上各種學習行為進行分析，找出各個行為在整體表現中的比重，進一步換算成學生線上表現分數，再進行各項關聯度分析及實驗，得到線上表現計分之分數與學習成就具關聯性。本研究將線上表現差的學生分為實驗組與對照組，對實驗組學生施以補教學習，發現實驗組在線上表現及學習成就，比對照組有明顯進步。
另一方面，本研究以學生合作學習分組策略研究為主，針對不同的分組策略在學習過程之中給予學生進行合作學習; 以知識互補、學習風格互補及思考風格互補作為合作學習分組策略，並對各種合作學習分組策略做比較，來觀察何種分組策略對學生的學習成效是有較大的幫助。實驗結果證明，影響學生合作學習之學習成就最主要的因素為學習動機，學習動機較佳的小組其學習成就也較佳。如果學生擁有良好的學習動機與知識互補再搭配其他風格互補分組，都能有效提昇學生之學習成就。各組學生在學習過程若發生學習動機低落或知識互補不足時，必須提供有效的補救措施，以提昇學生學習動機及知識互補，必要時更要予以重新分組。

關鍵字

學習動機 ； 合作分組策略 ； 主成分分析 ； 學習歷程 ； 適性化警訊 ； 集群分析

並列摘要

The diagnostic system proposed in this research is based on online learning portfolios. The system utilizes data mining, clustering analysis, and statistical tools for clustering students according to their online learning behaviors, learning achievements, and thinking styles. By using this system, teachers can apply different remedial strategies according to the different online learning behaviors of students. The students in this study received adaptive warning messages and remedial treatments according to their individual learning situations. The results revealed that the proposed strategy can encourage students whose online learning performances are initially evaluated as good or medium and improve their online learning performances. In the case of students with poor online learning performances, this system improves not only their performances but also their learning achievements. This is achieved by continuously dispatching adaptive warning messages to the students in question. The relationship between thinking styles and online learning achievements is also considered in this study. It is indicated that a strategy of sending adaptive warning messages also works well on students with different thinking styles.
Numerous learning behaviors are involved in this process. With so many factors involved, it can be quite difficult for instructors to assess the overall performance of learners. Therefore, it is desirable to develop a credible method of summarizing the online behaviors of learners, thereby producing, automatically, an integrated score for those behaviors. The resulting method can significantly assist instructors in evaluating the e-learning activities of their students. Principle component analysis was used to develop such a method. First, the various online behaviors of learners were analyzed, resulting in the determination of a weight for each item of measurement. Then, scores representing the online performance of the learners were computed. By correlating the computed scores to the learning achievements of the learners in question, it was found that the two share a significant relationship. Based on this finding, an experiment was conducted. In this experiment, learners with unsatisfactory online performance were divided into two groups. The experimental group received remedial treatments, whereas the control group did not receive any remedial treatment. The result was that the experimental group performed significantly better than the control group.
On the other hand, this study also proposes cooperative learning as remedial treatment, and explores the effects of different grouping strategies on cooperative learning. Learners were grouped using different strategies during the learning process; the grouping strategies used were based on the complementation of knowledge, learning-styles, and thinking styles. From observing and measuring the effect of a grouping strategy on overall learning performance, the three grouping strategies were contrasted. The results show that the two most important factors of cooperative learning are learner motivation and complementation of knowledge. Learner motivation plays an important role in learning. When motivated learners are teamed on the basis of complementation of knowledge, perhaps coupled with the complementation of learning styles or complementation of teaching styles, they generally learn better. it is important to give them timely remedial treatment to motivate them and to improve the complementation of knowledge. Sometimes, regrouping may be necessary.

並列關鍵字

Learning Portfolio ； Cluster Analysis ； Grouping Strategies for Cooperative Learning ； Principle Component Analysis ； Learner Motivation ； Adaptive Warning Messages

參考文獻

13. Chan, T. Y., Wang, R. M., Jong, B. S., Hsia, Y.T. & Lin, T.W. (2008). Conceptual graph-based learning material-producing strategy for cooperative learning. The 2008 IEEE Frontiers in Education Conference (FIE 2008), New York, USA, F2F15–F2F19.

12. Chen, C. M., Li, C. Y., Chan, T. Y., Jong, B. S., & Lin, T. W. (2007). Diagnosis of Students' Online Learning Portfolios,” The 2007 IEEE Frontiers in Education Conference, page: T3D-17-T3D-22, Wisconsin, USA.

38. Jong, B. S., Chen, C. M., Chan, T. Y., Hsia, Y. T.,& Lin, T. W. (2009).Applying Learning Portfolios and Thinking Styles to Adaptive Remedial Learning. Computer Applications In Engineering Education, Accepted.

2. Alfonseca, E., Carro, R.M., Martín, E., Ortigosa, A. & Paredes, P. (2006). The impact of learning styles on student grouping for collaborative learning: a case study. User Modeling and User-Adapted Interaction, 16(3–4), 377–401.

3. Borko, H., Michalee, P., Timmons, M., & Siddle, J. (1997). Student Teaching Portfolios: A Tool for Promoting Reflective Practice. Journal of Teacher Education, 48(5), 345-357.