How to effectively and efficiently represent the dynamics of spatial phenomena and processes has been a long-standing research question in geographic information science (GIScience). In a digital information age, computer-generated animations that depict movement data have become increasingly popular, as they apparently visualize real-world spatio-temporal movement changes with corresponding changes over time in a moving display. Animation thus seems to be a suitable display method for facilitating the recognition of spatio-temporal movement patterns and the prediction of future spatio-temporal events.
However, the manner by which animations are designed may limit the effectiveness and efficiency of visuospatial decision-making.
Furthermore, the specific decision-making task or context of use, as well as the viewer’s perceptual, cognitive and affective background might also influence visuospatial decision-making with animations. These factors are not well understood to date. More empirical studies, as well as new methods to evaluate animations, are thus needed.
This work proposes a user-centred empirical approach to evaluate animation design characteristics for space-time decision-making with movement data. Two experiments are conducted with the overall aim of answering the following main research question: How should animations of real-time movement data be designed considering the task and/or use contexts, and user characteristics? More specifically, we test the influence of the three main visual analytics (VA) dimensions on viewer spatio-temporal decision-making with animations: (1) the use context and respective task characteristics, (2) the animation display design, and (3) user characteristics. To test each respective dimension, we undertook the following investigations:
(1) Using current air traffic control (ATC) scenarios and existing ATC displays we empirically investigated how aircraft movement changes and future aircraft movement patterns can be visualized for effective and efficient decision-making in ATC. (2) We empirically investigated how movement characteristics (i.e., acceleration, heading direction, etc.) can be depicted, and how animation design (i.e., continuous vs. semi-static animations) might influence viewer task performances.
(3) We empirically investigated how perceptual, cognitive, and affective characteristics of viewers (i.e., expertise, spatial abilities, stress or motivation) might influence visuospatial decision-making with animations.
We approached these questions through novel empirical data triangulation that integrates psychophysical sensing (i.e., electrodermal responses (EDA)), brain activity (i.e., electroencephalography (EEG)), and eye tracking (ET) with standardized questionnaires.
The results of the experiments showed that these three factors (i.e., the use context and respective task characteristics, the animation display design, and the user characteristics) indeed influence visuospatial decision-making using animations of aircraft movement data. We found that viewer decision-making was affected by animation design depending on expertise and task type. Unsurprisingly, ATC experts performed typical ATC tasks more accurately compared to novices. However, the task performance of the experts differed between continuous animation and semi-static animation designs depending on the ATC task. Surprisingly, experts responded more accurately with the novel continuous animation designs compared to the semi-static animations that are more familiar to them in critical ATC tasks for predicting future aircraft movements. In apprehension tasks of aircraft movement changes, experts performed in similar ways with both animation designs. Moreover, viewer characteristics, such as spatial abilities and emotional aspects including engagement or motivation, seemed to affect viewer task performances as well. Higher-spatial and more engaged (or more motivated) viewers performed both tasks more effectively than lower-spatial decision makers and less-engaged (or less-motivated) viewers.
Overall, our unique empirical results related to the depiction of real-time movement data contribute to GIScience and cartography in two important ways. First, we are beginning to better understand how viewer mental processes, including perception and cognition, as well as their affective states might influence the effectiveness and efficiency of visuospatial decision-making with animations. Second, we are now able to derive empirically validated design guidelines for perceptually salient, affectively engaging, and cognitively inspired animations.