Characterizing joint attention dynamics during collaborative problem-solving in an immersive astronomy simulation
Characterizing joint attention dynamics during collaborative problem-solving in an immersive astronomy simulation
复制标题
DOI：
--
复制
发表时间：
2022
期刊：
影响因子：
--
通讯作者：
Yiqiu Zhou;Jina Kang
中科院分区：
文献类型：
--
作者：
Yiqiu Zhou;Jina Kang
协作的复杂和动态性质使得寻找生产性学习和高质量协作的指标具有挑战性。这项探索性研究基于学生使用增强现实耳机和平板电脑与身临其境的天文模拟交互时生成的日志文件，开发了一种协作指标来捕获联合注意(JA)的时间模式。JA被定义为协调注意力的能力，因此在协作解决问题以建立知识共建的共同点方面发挥着重要作用。我们首先开发了由六个不同但密切相关的状态组成的JA度量，作为协作过程的度量。然后，我们进行了描述性统计，以比较三个学习表现组中JA状态的频率和时间模式。我们的结果表明，高学习获得组表现出更频繁的视觉协调行为，并在早期阶段使用这种协作策略。然后，我们研究了这些JA状态的序列，重点关注一个关键行为：作为协作代理的长期且一致的共享视图。这一顺序分析揭示了两种不同的协作模式：注意跟随者和轮换者，这表明存在不对称参与。我们的发现表明，JA指标具有预测总体协作质量、识别不良协作行为以及作为提供即时指导的早期预警的潜力。
The complex and dynamic nature of collaboration makes it challenging to find indicators of productive learning and quality collaboration. This exploratory study developed a collaboration metric to capture temporal patterns of joint attention (JA) based on log files generated as students interacted with an immersive astronomy simulation using augmented reality headsets and tablets. JA is defined as the ability to coordinate attention, which thus plays an important role in collaborative problem-solving to build the common ground for knowledge co-construction. We first developed a JA metric consisting of six distinct but closely relevant states as a measure of the collaboration process. We then conducted descriptive statistics to compare frequency and temporal pattern of JA states across three learning performance groups. Our results showed that high-learning-gain groups demonstrated visual coordination behaviors more frequently and utilized this collaboration strategy in the early stage. We then investigated sequences of these JA states, focusing on one key behavior: long and consistent shared view as a proxy for collaboration. This sequential analysis revealed two different collaboration profiles: attention follow-leader and turn takers, suggesting the existence of asymmetrical participation. Our findings indicate the potential of JA metric to predict overall collaboration quality, identify undesirable collaboration behaviors, and serve as an early warning to provide just-in-time guidance.