Abstract
Dynamic data visualizations can convey large amounts of information over time, such as using motion to depict changes in data values for multiple entities. Such dynamic displays put a demand on our visual processing capacities, yet our perception of motion is limited. Several techniques have been shown to improve the processing of dynamic displays. Staging the animation to sequentially show steps in a transition and tracing object movement by displaying trajectory histories can improve processing by reducing the cognitive load. In this paper, We examine the effectiveness of staging and tracing in dynamic displays. We showed participants animated line charts depicting the movements of lines and asked them to identify the line with the highest mean and variance. We manipulated the animation to display the lines with or without staging, tracing and history, and compared the results to a static chart as a control. Results showed that tracing and staging are preferred by participants, and improve their performance in mean and variance tasks respectively. They also preferred display time 3 times shorter when staging is used. Also, encoding animation speed with mean and variance in congruent tasks is associated with higher accuracy. These findings help inform real-world best practices for building dynamic displays.
Introduction
We live in an era of "streaming" data, where real-time monitoring of heart rates, network traffic, or stock prices is commonplace. Visualizing this data often involves dynamic displays where lines or dots move across a screen. However, human perception has a significant "bandwidth" problem: we typically struggle to track more than four moving objects simultaneously. Furthermore, the transient nature of animation means data often disappears before it can be fully processed. This paper addresses the gap between the popularity of dynamic visualizations and our limited perceptual capacity, asking how motion—through staging, tracing, and speed—can be manipulated to make dynamic charts as effective as static ones.
Methods and Design Space
The researchers explored a design space defined by three primary variables: Staging (sequential vs. synchronous display), Tracing (showing the line path vs. a moving dot), and History (retaining the line on screen after the animation ends).
Through four experiments involving nearly 400 participants, the study tested these design options using both synthetic data (7,560 unique visualizations) and real-world Content Delivery Network (CDN) latency data provided by Dolby Laboratories. The experiments covered design combinations, speed preferences, speed as a redundant encoding (e.g., faster movement representing higher values), and real-world validation.
Key Insights
The Power of History
The most significant finding was the impact of History. Preserving the path of a line after the animation finishes was the single most effective way to boost accuracy, improving performance by 1.36 times. This technique effectively mitigates the "fleeting" nature of animation by turning a dynamic task into a semi-static one.
Staging and Speed
Participants preferred "staged" (sequential) animations to move significantly faster—up to three times faster—than synchronous ones. This suggests that when the visual system focuses on one entity at a time, it can process much higher velocities without a loss in comprehension.
Congruency in Encoding
Using animation speed to encode data (e.g., faster = higher variance) is effective only when it is congruent with the user's task. If speed encodes the mean but the user is asked about variance, performance drops as the motion becomes a distraction. When explicitly informed of the encoding (e.g., "faster = higher value"), performance in variance tasks improved by 1.31 times.
The Preference-Performance Gap
A critical takeaway for designers is the "asymmetry" between what users like and how they perform. Participants often preferred static charts or specific animation styles that did not yield their highest accuracy.
Conclusion
While static charts remain a strong baseline, well-designed dynamic visualizations can be highly effective for specific analytical tasks. Outlier detection, in particular, was highly accurate across all styles, suggesting human vision is naturally tuned to catch sudden spikes in motion. For more complex tasks like identifying mean or variance, incorporating data history and ensuring congruent speed encoding are vital. Future work should explore how these techniques scale to information-rich environments like multi-chart dashboards and long-term monitoring scenarios.
Summary updated