The art and practice of visualizing data is becoming ever more important in bridging the human-computer gap to mediate analytical insight in a meaningful way.
Data Visualization (or “Data Viz”, for short) has become a hot topic in the marketing research and analytics worlds of late. The folks at SAS define the concept as, “…the presentation of data in a pictorial or graphical format. For centuries, people have depended on visual representations such as charts and maps to understand information more easily and quickly.” 1 I believe that last part is the key, “to understand information more easily and quickly”. I want to take a few minutes to look at two applications we have as researchers to incorporate better data visualizations into our work.
Before we jump in though, a brief aside about the comment that these techniques have been around “for centuries”. Lest you think this is just something that is now possible due to the advent of modern computing capacity, speed or accessibility, take a look at the graphic below. It was created in 1869 by Charles Minard to depict Napoleon’s Russian Campaign in 1812-13. While advancements in the last 25 years have made it far easier to come up with certain visualizations, the concept itself has been around for a long time.
(For a detailed discussion of what’s going on in this chart, visit //tinyurl.com/o2s6urn.)
The first area of application for Data Viz concerns our internal ability to understand what story the data is trying to tell us. As researchers, we utilize a number of Exploratory Data Analysis (EDA) approaches to understand the big picture story that is lurking behind the data. By employing some creative ways of viewing the data, we can help ourselves by speeding up the time it takes us to find the story.
Take a simple correlation matrix, for example. Traditionally, we would look at something that looks like the right side of the graphic below. We would search through the matrix to determine which attributes are most correlated with each other. The raw output (in this case from SPSS) is just not that user friendly in terms of quickly understanding what might be going on. Contrast this with the left hand side of the graphic. This is output from a module in the statistics software package R that automatically color codes, sorts, highlights significance and generally makes it much easier to start understanding any patterns that may exist and any potential clues to the underlying story in the data. (Hat tip to Dr. Rob Kabacoff and his presentation on R and Data Visualization at the ASA CSP conference for the left side of this graphic.)
The second area of application for Data Viz is to aid in our story-telling ability. Recently we have moved away from just providing data to our clients to a more consultative place of finding the story within the data. Better data visualization techniques and incorporation into our reporting will allow us to tell the story to our clients more effectively.
As we do this though, I would like to make a few suggestions to help guide this pursuit:
1. Remember the audience.
There are some clients that may get turned off by, what they may perceive as, “oversimplifying” the data and the story. Alternatively, there are others that would embrace the idea of having a less rigid approach to the reporting as we incorporate better data viz representations of the story.
2. Don’t lose sight of the context.
It can be very easy to come up with some creative way to display something that is in no way related to the topic at hand. For example, if we are researching the brand equity of a chicken restaurant, it probably doesn’t make sense to use pie graphs that are made to look like pizzas!
3. Don’t get carried away.
This isn’t about turning hard core data into pretty pictures. There can be a fine line between insightful, creative displays of an idea and something that looks like it came out of a kid’s comic book. Remember that the idea is to help the reader (client) more efficiently and effectively understand the point of the story that we are telling through the data.
In the coming weeks and months I challenge us all to do a better job with both of these applications. There will be more discussion of this topic and we’ll provide forum(s) to share any ideas that you may come up with. Have fun with it! Just remember, ultimately the objective is to be both more efficient and more effective, at what we do.
1 Source: //www.sas.com/en_us/insights/big-data/data-visualization.html