CHS: Small: Collaborative Research: Validating and Communiciating Model-Based Approaches for Data Visualization Ability Assessment

人们现在比以往任何时候都更多地接触到数据的图形、图表和其他可视化表示。然而，这些可视化的创造者目前必须用稀少和相互矛盾的证据来论证人们对他们发布的可视化的阅读能力。目前的指南没有考虑到不同的人在解释视觉数据时有不同的优势和劣势的可能性。这个项目将使用可视化有效性的研究来帮助我们理解观众的能力和偏见，包括个人和集体。为此，项目团队将使用实验、统计建模和访谈研究的组合来挑战关于可视化效果的长期假设，并为未来解释可视化阅读能力差异的实验奠定基础。这项工作还将支持一个更广泛的教育目标，即在实验中使用稳健的统计建模技术，通过可以整合到现有数据可视化课程中的课程模块，以及通过外展活动，允许个人查看他们与其他参加过实验的人相比，他们执行可视化任务的情况。第一个是通过使用透明的统计方法来确定个人在数据可视化表现上的个体差异，通过大规模的众包实验，确定个人在执行数据可视化基本任务方面的差异程度。第二个问题评估低水平的可视化表现与更高水平的评估之间的关系，例如可视化素养和认知能力，招募专家和新手来评估这些假设的可视化素养测量相互关联的程度。第三个问题决定了呈现可视化实验结果的其他方法如何形成研究人员和设计者从中得出的设计建议，通过对长期以来呈现可视化实验结果的方法的比较评估，以及通过设计新的呈现结果的方法来可能导致更广泛的可视化社区对实验结果的更成熟的解释。这项工作将为可视化素养提供新的视角，通过使用新的可视化能力测量方法来增强海图阅读实验，并通过研究创建者目前如何在他们的设计过程中利用现有的可视化设计指南来提供新的视角。该奖项反映了NSF的法定使命，并通过使用基金会的智力优势和更广泛的影响审查标准进行评估，被认为值得支持。

People are encountering graphs, charts, and other visual representations of data now more than ever before. Yet creators of these visualizations currently must reason with sparse and conflicting evidence on how well people can read the visualizations they publish. Current guidelines do not take into account the possibility that different people have different strengths and weaknesses when interpreting visual data. This project will use studies of visualization effectiveness to inform our understanding of the abilities and biases of viewers, both individually and collectively. To do this, the project team will use a combination of experiments, statistical modeling, and interview studies to both challenge long-standing assumptions about visualization effectiveness, and to lay a foundation for future experiments that account for differences in visualization reading ability. The work will also support a broader educational goal of using robust statistical modeling techniques in experimentation, through course modules that can be integrated into existing data visualization courses, and through outreach activities that allow individuals to see how well they perform visualization tasks compared to others who have taken the experiments.This work seeks to answer three primary research questions. The first is to determine the extent to which individuals differ in their ability to perform basic tasks with data visualizations, through large-scale crowdsourced experiments that use transparent statistical methodologies to establish individual differences in data visualization performance. The second question evaluates the relationship between low-level visualization performance and higher-level assessments such as visualization literacy and cognitive abilities, recruiting both expert and novice populations to evaluate the extent to which these hypothesized measures of visualization literacy correlate with each other. The third question determines how alternative ways of presenting visualization experiment results shape the design recommendations researchers and designers draw from them, through a comparative evaluation of longstanding ways of presenting visualization experiment results, and by designing new ways of presenting results that may lead to more mature interpretation of experiment results by broader visualization community. The work will provide new perspectives on visualization literacy by augmenting chart reading experiments with novel measures of visualization ability, and by studying how creators currently make use of existing visualization design guidelines in their design process.This award reflects NSF's statutory mission and has been deemed worthy of support through evaluation using the Foundation's intellectual merit and broader impacts review criteria.