以網路同儕互評系統輔助高中電腦課程教學：學習成效及同儕回饋之分析

Abstract

學習的評量影響學生的學習成效，在多元化教學及注重多元智慧的學習環境中，傳統紙筆測驗的評量反應出學習的成效有限；同儕互評正是因應多元評量發展趨勢而生的學習模式，而網路更將這種同儕互評的學習模式發揮到淋漓盡致。在互評的過程中，每位學生分別扮演作者、評論家及修改者的角色，在角色切換的過程，不但可以讓學生主動學習或增強主動學習的意願，還增進學習者高層次具批判性的思考；藉由網路系統作媒介控制，還可消除了時間和地點的限制，並具有良好的匿名機制，讓傳統紙筆模式的同儕互評活動，更具公平、有效率。 本研究從建構主義的角度，探討同儕互評的理論基礎，並根據相關研究的網路同儕互評模型的修正，設計建構本研究的網路同儕互評系統(3-Rounds Peer Assessment, 3RPA)，並使用此系統來進行網路同儕互評的教學活動。 本研究研究對象為高中四個班級共一百八十四人，以班級為單位，進行三回合共六週的網路同儕匿名互評活動。每位同學匿名評論十位同班的其他同學，所以，每一位同學每一回合會收到十篇其他同學給予的匿名評論與評分。在每一回合結束後，專家老師也一樣評量所有學生作品；因為本研究是同儕互評，學生在系統中將無法查看老師給予的評量。實施同儕互評的前後，本研究也對所有對象進行同儕互評的態度評量，以瞭解 學生在實施同儕互評前後的態度與看法。 研究結果發現，使用網路同儕互評，的確對於能改善學習者的作業品質，且與專家教師的評量有顯著相關，表示高中生的同儕評量可參考性頗高。另外，本研究重點之一，在分析同儕回饋對於互評結果的影響，將約五千五百多篇同儕回饋的評論予以分類，統計哪些種類的回饋，能夠增進同儕的學習品質，對作業改善具有幫助。研究發現鼓勵性回饋對於各評分向度都有正向影響，而教誨性回饋對於各評分向度有負向影響。對於高中同學在實施網路同儕互評的的前後態度上，並沒有很大的落差。 本研究在實施後測態度評量，讓研究對象填寫心得感想與建議，發現大多數的高中同學對於本研究網路同儕互評活動與使用的系統(3RPA)，皆給予肯定的感想與建議，也有同學提出寶貴的建議與使用系統上的缺點，本研究也將資料整理，希望為未來相關學者的研究提供參考與依據。 值得注意的是，本研究的同儕回饋分析結果顯示，教誨性回饋對於學生的學習有負向的影響力，而傳統的教學模式中，大部分教師給予學生的，卻也以教誨性回饋居多。所以筆者認為，網路同儕互評雖然不是學習的萬靈丹，卻能夠提供學生另類的學習管道。而網路同儕互評系統也確實能輔助學科教學，的確提升了學生的作業品質，是值得推廣的輔助教學模式。

Evaluation plays an important role in students’ learning. Under the environment of Multiple Intelligences teaching and learning, conventional paper-based evaluation shows more restriction in evaluation. Peer assessment finds its beginning with the development of multiple assessment while the network makes it more mature. During the process of peer assessment, each student functions as an author, a reviewer, and an adapter. During the process of role transfer, students will not only learn more actively and willingly, but also attain higher critical thinking skills. With the network system as a medium and control center, we can eliminate the restrictions of time and place. Moreover, the anonymity of networked environment will create more fairness and effectiveness than paper-based peer assessment does. This study first discussed the theoretical foundations of peer assessment from a constructivist perspective and then modified the previous networked peer assessment models. Finally, a networked peer assessment activity was conducted by designing a web-based three round peer assessment (3RPA) system. The subjects of this study included 184 tenth graders, each of them from four different classes in Hsinchu Senior High School. The 3-RPA activity was performed on a class basis for six weeks. In each round, every student was required to critique the work of other ten counterpeers anonymously. Thus, each of them was supposed to receive ten comments and scores from their peer reviewers. At its end, experts (e.g., teachers) would evaluate students’ work as well. However, the results of the evaluation were not accessible to any of the students. Most importantly, students’ attitudes toward networked peer assessment were collected before and after the activity. Some significant findings are revealed in this study. Firstly, the networked peer assessment actually can improve the quality of students’ work . Secondly, the scores of peer reviewers are highly consistent with experts’ grading, indicating the validity of senior high students’ peer assessment. Thirdly, reinforcing feedbacks have more positive influences on every aspect of students’ work while didactic feedbacks have negative influences on it. With regard to students’ attitudes toward the networked peer assessment, no significance was found between before and after the activity. This research also allowed all of the participants to write down their feelings and suggestions about the networked peer assessment and the system they used (e.g., 3RPA ) after the activity. Most of them listed the merits of the activity while some of them found out some limitations about the system. All of their feelings, praises and suggestions were compiled for the use of future research and references. As what is shown in the peer feedback analysis, didactic feedbacks have negative influence on students’ learning. However, in traditional classrooms, most teachers tend to provide their students with didactic feedbacks. Therefore, the author suggests that though networked peer assessment is not a panacea for learning, it offers students an alternative way while they are learning. It is true that with the networked peer assessment system as a learning aid, the quality of students’ work is improved. And that accounts for why networked peer assessment system is worth promoting.