论文标题
评估时间轴形状对可视化任务性能的影响
Evaluating the Effect of Timeline Shape on Visualization Task Performance
论文作者
论文摘要
时间表通常在水平线上表示，这不一定是可视化时间事件序列的最有效方法。但是，很少有实验评估时间表形状如何影响任务绩效。我们介绍了在亚马逊机械TURK（n = 192）上进行的受控实验的设计和结果，其中我们评估时间表形状如何影响任务完成时间，正确性和用户偏好。我们测试了4种形状的12个组合 - 水平线，垂直线，圆圈和螺旋 - 以及3种数据类型 - 复发，非循环和混合事件序列。我们找到了充分的证据表明，时间表形状有意义地影响用户任务完成时间，但不正确，并且用户具有强大的形状偏好。在结果的基础上，我们提出了根据用户任务和数据类型创建有效时间表可视化的设计指南。本文的免费副本，评估刺激和数据以及代码可在https://osf.io/qr5yu/上获得。
Timelines are commonly represented on a horizontal line, which is not necessarily the most effective way to visualize temporal event sequences. However, few experiments have evaluated how timeline shape influences task performance. We present the design and results of a controlled experiment run on Amazon Mechanical Turk (n=192) in which we evaluate how timeline shape affects task completion time, correctness, and user preference. We tested 12 combinations of 4 shapes -- horizontal line, vertical line, circle, and spiral -- and 3 data types -- recurrent, non-recurrent, and mixed event sequences. We found good evidence that timeline shape meaningfully affects user task completion time but not correctness and that users have a strong shape preference. Building on our results, we present design guidelines for creating effective timeline visualizations based on user task and data types. A free copy of this paper, the evaluation stimuli and data, and code are available at https://osf.io/qr5yu/
