網路同儕評量將教育評量網路化，兼具同儕評量及網路輔助教學的優點。本研究以自我調制學習理論觀點，探討網路同儕評量的特定環境下，學習者藉由社會互動以建構知識的過程，分析學習者的認知處理、回饋如何在學習中產生作用。希冀鼓勵教學者瞭解回饋對學習者自我調制的交互影響。最後根據研究發現提出具體建議，供日後研究與教師從事教學之參考。 研究對象是北部某大學研究生共12位，科目為「教育與心理統計學」。研究採取個案方式進行，在學科內容教授完畢後，進行網路同儕評量活動，學習者完成學期作業，共歷時10週。蒐集資料以問卷、晤談、內容分析（作業內容、學習者自評、心得內容）、回饋內容分析為主。 研究結果如下：研究目的之一在於建構網路同儕評量自我調制學習模式。隨互評次數的多次歷程，學習者認知系統的知識的信念與動機、目標設定、學習策略實施、監控、知覺外在回饋、反省與評估等不斷持續進行的階段性過程所組成。在網路同儕評量回饋內容分析上，互評的次數的增加對於評審者給同儕高層次回饋功能的次數並沒有增進，但深入分析回饋內容，就大多數單一個案，其回饋內容卻更有品質。口頭與紙本回饋類型與回饋功能有顯著關連。回饋類型對回饋功能具有14.5%的預測力。學習者接受回饋後修正作業表現程度有顯著的效果。學習者於研究者自行編制的問卷「網路同儕評量回饋認知功能效果」有良好的回應。有91.7%的學生肯定網路同儕評量活動，願意再次參與。 Networked Peer Assessment adopts Internet resources to assist assessment. It has advantages of peer assessment and computer-assisted instruction. This study has applied networked peer assessment to explore the processes of interactive knowledge construction of learners. We have used self- regulated learning theory to analysis the cognitive processes of learners and how feedback functions and then constructed a model of self-regulated learning under networked peer assessment. We are looking forward to instructor to understand feedback and self-regulated learning of learner are reciprocal effect. Finally, suggestions for teacher instruction and future research topics are proposed. Twelve postgraduate students at a national university in north Taiwan served as subjects. The curriculum is Education and mental statistics. The research was a case study. After teaching content of academic subject, teacher carried out networked peer assessment and learners finished study assignments in term. It lasts 10 weeks altogether. Collect the survey materials include questionnaire method, interview, content analysis (study assignments、learners wrote what they have learned and self assessment). The results were as follows: study one of purpose is to establish model of self-regulated learning under networked peer assessment. Because of the number of times of networked peer assessment, the cognitive system of learners will going on continuously constantly phased course make up by the belief of knowledge、goal setting、learning strategies、monitor、aware external feedback、self-examination and evaluate. The content analysis of networked peer assessment showing increase of the number of times has not been promoted the assessors provides high level functions of feedback. Researcher go deep into analysis the content of feedback, in most of a single case, the functions of content but there is quality even more. The types of feedback (the oral feedback and the written feedback) to the functions of feedback have significant association. The types of feedback has the capability of the index of predicative association is 14.5% to the functions of feedback. Learners revised the study assignments have noticeable effect after received feedbacks. The results of learner replied to the questionnaire, the effect of cognitive functions of feedback under networked peer assessment, edited by researcher, is good for learning. About 91.7% of subjects approve the networked peer assessment and would like to participate again.
