中文摘要：

网络协作学习是促进学习者高阶认知、协作能力发展的有效方法。目前,网络协作学习分析多以个体学习者的学习结果为研究对象,难以实现对以小组为单位的协作学习过程全面、深入的认识。小组网络协作学习分析模型以建构主义学习理论和活动理论为基础,构建了以主体、客体和共同体作为参与协作学习的主要对象,以工具、规则和分工作为影响协作学习过程与结果主要因素的分析模型,从小组中个体的贡献和小组综合状态两个维度,确定了13个可以显式观察的评价指标,实现了对指标的量化操作方法。该分析模型在研究生网络课程＂教育技术研究方法＂的应用中发现以下特征：学习者对小组任务的贡献方式与学习风格的异质性密切相关;协作学习并不因为创建了环境、规定了任务而必然发生;实时的、连续的分析反馈是引导网络协作学习深化发展的重要保障。实践证明,该分析模型不仅能够成为引导我们全面认识网络协作学习活动的有效框架,而且在指导网络协作学习分析实践方面具有良好的可操作性。

英文摘要：

Web-Based Collaborative Learning（WBCL） is an effective way to promote the development of learners＇ higher cognitive ability and cooperation ability. In present, most analysis of the web-based collaborative learning effectiveness takes the learning outcome of individual learners as the main object of studies, difficult to achieve a comprehensive and thorough understanding of the group collaborative learning process. In this paper, we constructed a WBCL analysis model based on the Constructivist Learning Theory and Activity Theory. The model takes the subject, object and community as the main objects of participating in collaborative learning and takes the tools, regulation and division of work as the key factors which will affect the process and result of collaborative learning. In addition, 13 evaluation indexes that can be significantly observed were identified from two dimensions including individual contributions in groups and group comprehensive state for quantitative operation. Finally, the graduate students＇ online course ＂Research Methods on Educational Technology＂ were analyzed through the WBCL analysis model, and the characteristics of WBCL in this course are concluded as follows： the learner＇s contribution to the group task is closely related to the learner＇s learning style; collaborative learning does not necessarily occur when the learning environment and the task are set; real-time and continuous feedback is an important guarantee for guiding the development of collaborative learning. Practice has proved that the model will not only be the effective framework to guide us in understanding the WBCL activities fully, but also has a good operability in supporting WBCL analysis.