Attention (or Lack Thereof): Understanding Limits of Visual Attention Span
How to design visualizations that will not overwhelm the viewer.
The design of an information visualization plays an important role in how the information is processed by the viewer. Components that can make up a visualization can include — but are not limited to — position, color, size, motion, and texture. Designers need to keep in mind the speed and capacity of human attention in order to create the best possible visuals.
Researchers at the UC Davis and UC Berkeley conducted three experiments (find a target, find the out of place object, and complexity assessment) studying how the combination of different visualization components to better understand these capacity limits. Each experiment examined how the viewer’s response time and accuracy changed as color and motion were altered. Colors or motions were grouped together by similarity or randomized in the visuals presented to the viewers.
In the first experiment the viewer had to search for a known target. Viewers were presented with a target to look for in a visual presented to them. During the second experiment the viewers had to identify the part of the visual that stood out from the rest. In the final experiment, viewers needed to quickly summarize or estimate the amount of variability in a visualization. These studies found that the viewers had an easier time completing all of the above tasks when similar color and motions were grouped together rather than randomized.
While these results seem like they would be common sense (that finding a target in a randomly arranged image is difficult), it is easy to find visualizations that take human visual capacity for granted.
The researchers came up with 5 information visualization guidelines based on their results.
1. Grouping similar data can aid in some visualization tasks
Use grouping to help make visualizations that cannot be done quickly just by viewing raw data such as identifying outliers and trends in the information. Visual features used to accomplish this goal need to be balance so that they do not impact performance. A visualization that is too colorful can make it more difficult to see the differences in a data set.
2. Change the task if data cannot be grouped
Letting the viewer know what to look for in advance through the use of a legend, key, or other label greatly improves task performance. These labels help the viewer perform a guided search through the data.
3. Less (categories) is more
Avoid presenting all information categories to the viewer at once. Limiting the information makes the visual easier to search through. Add options to allow the viewer to add or remove categories that they are interested in examining.
4. Assign visualization features to specific aspects of the data
Visualization features need to be chosen carefully. For example motion can be used to guide the viewers’ eyes, but it could distract from reading the information in the visualization.
5. Test and retest the visualization
Self-test your visualization and have others test it as well. See if your visual makes sense. Ask yourself is the presentation of the data effectively guiding the user, can this data be grouped differently, is the motion or color distracting?
These five simple guidelines will aid in the development of more impactful visualizations. They provide easy concepts to follow to avoid overwhelming the viewer and can be implemented in any data visualization tool.
This article is inspired by the following paper:
- Haroz, S., & Whitney, D. (2012). How Capacity Limits of Attention Influence Information Visualization Effectiveness. IEEE Transactions on Visualization and Computer Graphics, 18(12), 2402–2410. https://doi.org/10.1109/TVCG.2012.233