Skip to main content

Why reading charts is hard & How language and shapes can help

Charts and bar graphs exist within a visual space and rely on the human visual system to extract relationships from the 2D information. The information extracted pertains to how the variables within the 2D system relate to each other on a visual plane. Reading a graph isn’t as cognitively simple as it may seem. A graph with three bars requires a viewer to consider multiple pieces of information and reach a conclusion.

smiling man with glasses

Steven Franconeri, Northwestern Faculty and SILC Advisory Board Member

“In order to do that visually you have to do some visual spatial gymnastics in your head, where you have to imagine the average of just these two bars versus those two bars or the delta between these bars versus those bars. And that’s really tough to learn” says Steven Franconeri, a SILC board member and head of the Visual Thinking LabFranconeri conducts visual spatial research and studies how people process visual information. Reading a graph can be easier if we look at the graph in terms of shape.   

 Those who are more versed in statistics or accustomed to reading 2D visualizations of data may often use shapes to understand the relationships between variables on a chart. “The 2D visual spatial system lets you see patterns that are tough to see before. 2D patterns let you see relationships as shapes” said Franconeri. No one fully understands how or why some people are naturally better at visually extracting information from 2D systems. Yet there are simple “tricks” which can make reading graphs easier. For example, loops in a scatterplot graph, “are the result of similar patterns that are shifted by up to a quarter of the periodicity of the pattern” (Haroz et al., 2015). However, there are limits to how much the human brain can visualize even in people who are really good at doing this. The work of Franconeri provides insight into what’s going on in the brains of people who are naturally able to perform complex visual spatial tasks to apply this knowledge in real world contexts.

a looped graph

“Known concepts have new visual features” (Haroz et al., 2015)

The human brain perceives simple reversals as very different sentences. “If I said I was at a party last weekend and said I met Hillary Clinton versus if I said I was at a party last weekend and Hillary Clinton met me.” Even though both sentences deliver the same information about the main event, you know there is a difference. This simple example illustrates how important it is to coordinate language with the visual space a viewer is asked to consider.  “We try to show that the way that language guides your attention helps you pull spatial relationships out of the world.” Groups or people interested in data visualization, such as journalists, academics, or students, can leverage the spatial relationships cued by language and shapes to better understand information presented as charts or via text.  

 

Works Cited: 

 Haroz, S., Kosara, R., & Franconeri, S. L. (2015). The connected scatterplot for presenting paired time series. IEEE transactions on visualization and computer graphics22(9), 2174-2186.