Data Viz: Are You Telling a Story or Making a Point?

April 17, 2017 10:20 am PST | Open Data

“How many people have ever said to themselves or a colleague or a client, let’s tell stories with these data?” asked Jonathan Schwabish, Senior Fellow at the Urban Institute and founder of the data visualization firm PolicyViz. “I’m going to make the case to you…that we’re not telling stories.” At the Socrata Connect conference in Washington D.C., Schwabish, author of “Better Presentations: A Guide for Scholars, Researchers, and Wonks,” argued that most data analysts are not telling stories with data, but instead making a point.

To illustrate his case, Schwabish told an anecdote about a member of the Council of Economic Advisors who put together charts illustrating different facets of life on Indian Reservations. These charts spurred then President Obama to visit tribes in Dakotas where he and the First Lady heard firsthand accounts from teens about the hardships of growing up on a reservation. Only after hearing these stories, and not upon seeing the charts, did funding different programs on reservations become a priority for the administration. So what made the stories, and not the data showing the same information, effect change?

 

What Defines a Story

Schwabish says stories include two key factors:

  1. Emotion: Something that makes you feel viscerally, not just a feeling like angry or sad or happy, but a physical gut reaction or call to action
  1. Meaningful Climax: Not an ending per se, but a driving sense towards an emotional height

Schwabish then explained several story archetypes and how data analysts can use them.

 

 

Schwabish quoted the book, “The Seven Basic Plots,” by Chris Booker and shared the seven plots lines that make up every story. They are:

  1. Overcoming the monster
  2. Rebirth
  3. Rags to riches
  4. Voyage and return
  5. Comedy
  6. Tragedy
  7. The Quest

According to Schwabish, every story has at least one of these basic plots, while many stories contain two or more. Schwabish gives the example of “The Lord of the Rings,” which contains the major plot point of “The Quest,” but also includes voyage and return, overcoming the monster, and rebirth. He suggested to his audience that they figure out which plot points can be presented when looking at data. Then, they can apply a story structure.

 

3 Story Structures 

The first structure Schwabish presented was “The Hero’s Journey,” so called by Joseph Campbell, where the protagonist travels from the Ordinary World to the Special World following designated stops along the way. Flowing through all the points along the circle allows the reader to follow the narrative of changing the status quo, complete with challenges met along the way and how those trials were overcome, and make an emotional connection to the story being presented.

The second structure is the traditional story arc. In a clip from author Kurt Vonnegut, he explains that action starts along the y-axis of “very bad” to “very good” and then action happens along the x-axis or “beginning” to “end.” How the story arcs along those axes is up to the writer.

The third structure is known as Freytag’s Pyramid, and that might be remembered from middle school. The hero is introduced, she has challenges leading to the height of action and emotion, and then goes back to a resolution.

 

How All This Pairs with Data 

All of the above points can boil down to three basic takeaways for the data analyst:

  1. Annotation: Add things to the graph to help the user understand. Give the reader an idea how the chart is to be read and add needed context.
  2. Narration: Add things to guide the reader through a single graph or a series to help them better understand the point being made.
  3. Story: Find the emotion or meaningful climax in what is being presented. “It’s the stories from the individuals that really makes us care,” quotes Schwabish from journalist Sarah Kliff.

 

 

In a perfect of example of facts meeting story, Schwabish directs the audience to a piece called “Snowfall” that originally ran in the New York Times. The story details a fatal avalanche in Washington State in 2012. The piece not only combines data about avalanche facts and statistics, but ties in the emotion felt by those involved with direct quotes and chilling firsthand accounts.

 

Putting It All Together

“It’s about pairing the analysis with the people, so that our reader and our user and our audience care about the content that we are conveying to them,” concluded Schwabish. Want to learn more? Check out Jon Schwabish’s full talk in the video below and you can find his new book,”Better Presentations: A Guide for Scholars, Researchers, and Wonks,” along with more resources and downloadable materials at his website, PolicyViz.