How Should I Interpret a Data Visualization?
Data visualizations can take on multiple formats and can represent a diversity of information types and combinations, all of which can impact your ability to understand what is being represented. According to Justin Joque in a chapter from Creating Data Literate Students (2017), "data visualization allows us to see something—for example, patterns, trends, or anomalies—in the data that we otherwise would not see."
As a reader, your goal is to understand, interpret, and reflect on the information represented in a data visualization and then infer conclusions based on that assessment. Let's use the visualization below to walk through some tips on how to approach this process.
Establish what the data visualization is trying to convey. The main aim of any visualization is to communicate or support a specific idea or message. It is up to you, the reader, to use the visualization and any accompanying narrative to determine what message the author is trying to communicate. For instance, the line chart above is measuring homelessness in the state of Montana and might aim to demonstrate that homeless populations are decreasing or increasing depending on the author's viewpoint.
Make explicit observations of the visualization. Quite literally, what do you see? Identify all key elements of the visualization, including the title and the labels presented on the x-axis (horizontal axis) and y-axis (vertical axis) of a graph or chart. Looking at the Montana chart again, we see that counts of homeless are tracked from 2007 through 2019.
What patterns can you discern? Depending on the visualization type, patterns can present themselves as clusters, upward (increases) or downward (decreases) trends, changing gradient shading across different map locations, and so on. Patterns like these are usually where the main upshot of the data visualization lies. In the Montana chart above, there is a clear consistent increase in numbers of homeless in Montana between 2009 and 2013 and an uneven decrease in the years afterward.
What other factors may have shaped the data and therefore the visualization? Think about what information is not included in the visualization: for instance, comparing homelessness in Montana to other states may not be appropriate if those other states use different definitions for what constitutes being "homeless." If we observe large gaps between each state's trend line, then the different "homeless" definitions by state may be contributing to those gaps. This type of information can change how we interpret a visualization and should be noted by its creator.
Reflect and interpret the information presented. Based on the patterns we can discern and factors that we know, does the upshot of the visualization support or undermine the author's message? For example, homelessness in Montana trended upward in 2009 through 2013 and has since decreased, but it doesn't necessarily mean that homelessness is no longer an issue in Montana.
Infer further. What other tentative conclusions can you deduce based on this interpretation? If homelessness in Montana increased 2009 through 2013, we might infer that this increase occurred as a result of the Great Recession based on our own knowledge of that time period. Be careful though—this is an inference that you would need to investigate and support with further evidence.
Not every data visualization will require a step-by-step thought process like this—the best visualizations are often the simplest and self-explanatory. However, it's always helpful to have an idea of where to start, especially if you're not comfortable with data or statistics. Nowadays, data visualizations are everywhere and because of that the ability to thoughtfully interpret them has become a critical skill to learn.
Reference
Joque, J. (2017). Making sense of data visualization. In K. Fontichiaro, J.A. Oehrli, & A. Lennex (Eds.), Creating data literate students. Michigan Publishing, University of Michigan Library, http://datalit.sites.uofmhosting.net/books/book/.