Situating Big Data: Assessing Game-Based STEM Learning in Context

这个真正的项目源于2013年关于改善教与学的数据密集型研究的征集。这一努力的目的是将来自不同学科的研究人员聚集在一起，促进使用与教育相关的大型数据集中的数据的新的、变革性的、多学科的方法，以创造可操作的知识，以便在中期改善STEM教学和学习环境，并在较长期内彻底改变学习。项目组的目标是了解如何使用从使用学习技术的环境中收集的数据来做以下工作：(1)允许进行自动化评估，在提供定制反馈建议时考虑到所有课堂活动和围绕技术使用的讨论；(2)更好地了解学习和学习发生的背景是如何相互作用的；以及(3)为改进学习方法和活动提供基于理论和证据的建议。这将使教师更容易管理正在进行的评估，并使课堂活动在以学习者为中心、基于项目和探究驱动的学习环境中适应学习者的需求。该项目的结果将为使评估常态化、常规化和持续进行以及更全面地考虑学习活动奠定基础。反过来，这将为学习者提供更好的个性化以及持续的反馈和支撑。结果将增强对如何评估和培养不仅是学科学习，而且还包括性格、认同发展和长期参与的理解。PIS寻求将情境认知理论与大数据分析相结合。他们将探索如何将来自技术的点击流数据与描述技术使用环境的关键形式的多模式数据相结合，例如个人和群体话语(在线和室内)、个人和课程人工制品、课堂评估和学校表现，以生成数据驱动的方法，以：(1)了解在技术丰富的学习环境中发生的学习；(2)以建议适应和支架的方式评估个人在这些环境中的发展和需求；以及(3)调查情境认知。它们旨在更容易地管理正在进行的评估，并使课堂活动在以学习者为中心、基于项目和探究驱动的学习环境中适应学习者的需求。他们将研究如何(1)在评估学习和参与度时，能够考虑整个学习生态系统和通过它收集的数据，以及(2)确定在某种情况下什么是有效的，什么是不有效的，以促进学习。他们将展示当学习者参与实践和语篇丰富的学习活动时，学习生态中的有用数据位于哪里以及如何定位，以及如何使用这些数据来评估干预措施的有效性和影响。他们的计划包括将手工编码的定性数据中的重要模式与自动收集的数据模式进行匹配；这将使他们能够识别自动数据收集中的模式，这些模式可以用作理解、困惑、学习和参与等因素的指示器。

This REAL project arises from the 2013 solicitation on Data-intensive Research to Improve Teaching and Learning. The intention of that effort is to bring together researchers from across disciplines to foster novel, transformative, multidisciplinary approaches to using the data in large education-related data sets to create actionable knowledge for improving STEM teaching and learning environments in the medium term and to revolutionize learning in the longer term. The project team aims to understand how to use data collected from the environment in which learning technologies are used to do the following: (1) allow automated assessment that takes the full range of classroom activities and discussions around use of the technology into account in providing customized feedback recommendations; (2) come to better understand how learning and the context in which it is happening interact; and, (3) provide theory-informed and evidence-based advice for refining learning approaches and activities. This will make it easier for teachers to manage ongoing assessment and to adapt classroom activities to learners' needs in learner-centered, project-based, and inquiry-driven learning environments. Results of this project will lay the foundations for making assessment regular, routine and ongoing and to take a fuller range of learning activities into account. This, in turn, will allow better personalization and ongoing feedback and scaffolding for learners. Results will enhance understanding of how to assess and foster not only disciplinary learning, but also disposition, identity development, and long-term participation.The PIs seek to integrate theories of situated cognition with analysis of big data. They will explore how to integrate clickstream data from technology with key forms of multimodal data describing the contexts in which the technology is being used, e.g., individual and group discourse (online and in-room), individual and curricular artifacts, classroom assessments, and school performance, to generate a data-driven methodology for: (1) understanding the learning happening in technology-rich learning environments; (2) assessing development and needs of individuals within those environments in ways that will suggest adaptations and scaffolding; and (3) investigating situated cognition. They aim to make it easier to manage ongoing assessment and to adapt classroom activities to learners' needs in learner-centered, project-based, and inquiry-driven learning environments. They will investigate how to (1) enable consideration of the full ecosystem of learning and data collected across it when assessing learning and engagement, and (2) identify what is working and not working to foster learning in a situation. They will demonstrate where and how useful data are situated in the learning ecology when learners are engaged in hands-on and discourse-rich learning activities, and how to use these data to assess effectiveness and impact of interventions. Their plan involves matching important patterns in hand-coded qualitative data to patterns of automatically collected data; this will allow them to identify the patterns in automated data collection that can be used as indicators of factors such as understanding, confusion, learning, and participation.