理論引導學習者中心訂正學習系統之精進工程：教育潛能初探性評估

Theory-Guided Learner-Centered Error-Correction Learning System Enhancement: Preliminary Evaluation of Its Educational Potential

有鑑於錯題訂正對學習的眾多效益，研究者甫成功建置一個符合永續學習概念的「學習者中心錯題訂正學習系統」。為進一步精進系統功能，本研究鎖定具理論基礎，惟尚待實證研究強化的三項設計（參與者學習進度資訊公開、求救對象選擇以及線上個人身分隱藏），進行教育潛能之初探性實證研究。研究邀集一所國小六年級四個班級學生（N = 104），配合其社會科進行10週教學，以評估學生對三項精進設計是否有顯著偏好差異，並瞭解個人偏好的考量點。主要結果為：一、卡方檢定結果顯示，大多數參與者支持成員學習進度資訊公開與能對求救對象進行選擇的精進設計；然而，對於隱藏成員身分之設計則有顯著較多參與者持反對立場。二、姓名顯示方式質性與量化分析結果凸顯社會因素對個人行為的影響力，並指出學業成就與朋友情誼為社會影響力的重要來源。三、本研究有近四分之一參與者在不同情境間會改變個人對匿名性之偏好立場，顯示學生立場的情境依賴性。最後，學生的詮釋揭示本研究三項精進設計能提供具行動指引性的有意義訊息內容與無助於任務完成的情緒負擔，進而對學習與任務完成有促進作用。文末，在提出有關系統設計與教學實施之建議，以營造更具效能、友善的錯題訂正線上學習環境與活動之後，說明本研究限制及具學術潛能的未來研發方向。

Considering the numerous benefits of error correction for learning, this study developed a learner-centered error correction system aligned with the principles of sustainable learning. To further improve the system, three features based on previous research leveraging the affordances of digital technology were incorporated: (1) shared information on participants’ work progress, (2) the ability to choose a help provider, and (3) identity concealment. Empirical evidence regarding these three features remains limited because previous studies have mainly involved adolescents or older participants with small sample sizes in short study durations and have resided mostly in media communication or interpersonal relationship fields. Thus, a 10-week study involving four purposively sampled sixth-grade classes (N = 104) was conducted to examine students’ preferences for different feature arrangements and the reasons underlying those preferences. Several notable findings were obtained. Firstly, chi-square tests revealed that a significantly high proportion of participants supported (1) access to peers’ work progress and (2) the ability to choose a help provider. However, this pattern was not observed for the enhanced design of identity concealment. Secondly, quantitative and qualitative findings regarding the participants’ revealed preferences indicated strong social influence from high-achieving peers and friendship ties. Thirdly, in situations involving anonymity, nearly one-fourth of the participants changed their selections across the three examined conditions, indicating that the participants’ preferences regarding anonymity were situation-contingent. Finally, the participants’ explanations for preferring a particular arrangement highlighted the distinct, enabling effects of the three enhanced features in supporting purposive follow-up actions and reducing emotional discomfort to uphold meaningful learning and successful task completion. This study provides suggestions for system design and instructional implementation to create a supportive online learning environment that enhances productive error correction activities. The limitations of this study and recommendations for future research are also discussed.