【作 者】：
姚佳佳，李 艳，刘明月，马志强
【关 键 词 】：
高校教学； 可视化实时协作； 协作问题解决； 群体深度学习； 多维特征挖掘
【栏 目】：
课程与教学
【中文摘要】：
随着对教育互动的研究越来越强调“情境依存性”，关注复杂交互场景下互动数据的多维分析有助于还原群体学习复杂过程、揭示特定规律、优化特定服务。研究通过对三组不同协作成效的个案在可视化实时协作场景中的互动数据进行协作关系、协作行为和协作时序的多维分析发现，群体深度学习发生的特征表现为：（1）有多个核心成员且分层领导的组织关系，更有利于促使小组在协作前期的资料搜集和推进方案想法或成果形成的阶段积极展开交互反馈；（2）小组在协作过程中基于“内容建构”的集中性反馈和基于“案例共享、信息搜集、想法拓展”的实质性反馈对小组协作深度影响大于围绕“线索提示、任务明晰”的流程性反馈；（3）小组协作习惯主要存在“问题导向、资料导向、任务导向”三种典型模式，对小组协作深度的促进依次减弱。该研究为高校教学多元交互场景中新协作样态下的群体深度学习发生机理揭示、评价方法和促进策略优化等提供了理论与实践基础以及后续建议。
【英文摘要】：
With the increasing emphasis on "context dependency" in the study of educational interactions, multidimensional analysis of interaction data in complex interaction scenarios can help to restore the complex process of group learning, reveal specific patterns, and optimize specific services. Through the multidimensional analysis of interaction data of three groups with different collaborative performance in visualized real-time interaction scenarios in terms of collaborative relationships, collaborative behaviors and collaborative time sequences, it is found that the features of group deep learning are as follows: (1) the organizational relationship with multiple core members and hierarchical leadership can be more conducive to encouraging the group to gather information and advance program ideas in the early stages of collaboration, or actively carry out interactive feedback in the stage of outcome formation;(2) in the process of group collaboration, the centralized feedback based on "content construction" and the substantive feedback based on "case sharing, information gathering and idea development" have a greater impact on the depth of group cooperation than the procedural feedback based on "clue prompts and task clarification"; (3) there are three typical modes of group cooperation habits, namely the problem-oriented, the data-oriented and the task-oriented, with the promotion of the depth of group collaboration being weakened in turn. This study provides a theoretical and practical foundation as well as subsequent suggestions for the revelation of the mechanism, the optimization of evaluation methods and promotion strategies under the new collaborative paradigm in the multidimensional interaction scenarios of higher education.