AI 中文总结
该研究开发了两种基于网络的可视化素养质性评估，与CALVI、Mini-VLAT对比后发现，其能捕捉独特技能且更好区分有经验个体，可作为现有评估的补充。
AI 中文摘要
可视化素养是一个多维度的概念，涵盖解读数据、构建图表、识别设计缺陷等技能与能力。然而，对这些能力的评估主要局限于针对图表理解等低阶技能的选择题评估，导致评估常出现天花板效应（即技能中等的个体也常获得接近满分的成绩），且无法提供足够信息来衡量个体的高阶技能（如应用外部知识、提出批判、设计可视化）。为填补这些缺口，我们开发并研究了两种基于网络的质性评估，通过在线出声思考批判、基于数据和提示绘制可视化设计的方式，测试可视化素养的批判与设计维度。我们将这些评估的表现与两种已有的可视化素养评估工具CALVI和Mini-VLAT进行对比，对代表三种经验水平的三组人群（众包工作者、修过相关课程的学生、研究人员）进行评估。研究发现，我们的批判和绘图评估捕捉到了与现有评估不同的技能，且比基于选择题的评估更能区分有经验的个体。尽管实施和评分质性评估存在挑战，但研究结果表明，质性多模态评估是现有可视化素养评估的有前景补充，尤其在需要区分高可视化技能时。
英文摘要
Visualization literacy is a multifaceted construct encompassing skills and competencies, such as decoding data, constructing charts, and identifying design flaws. Yet, assessments of these competencies has been primarily constrained to multiple choice assessments that target lower-order skills, such as chart comprehension. As a result, they often exhibit ceiling effects (i.e., even modestly skilled individuals commonly score near the top of the scale), and do not provide enough information about an individual's higher-order skills (e.g., applying external knowledge, formulating critiques, and designing visualizations). To close these gaps, we develop and investigate two web-based qualitative assessments for testing the critique and design aspects of visualization literacy through online think-aloud critique and sketching of visualization designs based on data and a prompt. We compare performance on our assessments to two established visualization literacy assessments, CALVI and Mini-VLAT, by administering them to three groups that represent three experience levels: crowdworkers, students who have taken a relevant course, and researchers. We find that our critique and sketching assessments capture skills distinct from existing measures and that they differentiate between experienced individuals better than multiple choice-based alternatives. Although administering and grading qualitative assessments can be challenging, our findings suggest qualitative, multimodal assessments are a promising complement to existing visualization literacy assessments, in particular when high visualization skills need to be distinguished.
