小組合作學習策略之研究

Abstract

合作學習是近幾十年來相當受到重視的學習方法，但影響合作成效的因素相當的多，本篇論文將針對分組分法和思考風格兩個主軸探討合作學習策略。 小組合作在各領域中都相當重要，但如何在有限的資源下做最有效的分組，目前多靠經驗和感覺，本論文設計了一個可以同時考慮多維變數的分組模型，並針對不同組隊動機採用不同分組演算法，以便應用所有資源能做有效分組，且整體的效能仍能維持住一定水準。這個分組模型和演算法在群組人力規劃和配置上將可兼顧平衡與效能兩方面的考慮。 為探討思考風格與個人及小組學習成效的關係，我們將實驗過程分為兩個階段，以有效掌握住個體和群體工作模式。利用網路問卷調查，取得學生本身原有的思考風格、個人的工作方式、該生在小組中的工作方式及互動模式。如此，我們便能觀察、驗證許多假設，並歸納出有價值的合作策略和互動方式。 本次實驗中，我們歸納出學生處理學習主題（人工智慧下棋程式）的策略有六種基本型態，而小組的合作模式可分為四種。這些分析方法均具有一般性，可以推廣至其他領域。

Team-based collaborative learning has become an important learning method in the past few decades. Many factors of learning effect have been discussed. In this study, we aim at developing a novel team-forming model, and investigating the strategies of collaborative learning based on the thinking styles of team members. In many fields, teamwork has been considered an essential part in success. But, how to partition given limited human resources? How to form teams? Most existing solutions depend on managers’ experiences and gut feeling. In this thesis, we design a team-forming model that takes multi-dimensional factors as input and adopts difference algorithms according to the user’s requirement. This model achieves a balance between team homogeneity and integral performance at the same time. To thoroughly investigate the correlation between thinking styles and team-based learning effect, we split an experiment into an individual stage and then a teamwork stage. Using several network questionnaires, we collect students’ thinking styles, individual working styles, team working styles, and team interactions. We analyze and discuss various correlations between them in this thesis. Based on this experiment, we induce six strategies that the students employed to cope with the leaning theme (artificial intelligence chess game), and find four kinds of collaboration style. These analytical methods provide a general framework and can be utilized in other fields.