5 Data Storytelling Homework Assignments

Level Up Your Data Storytelling Game

What draws me to data storytelling? The opportunity to work in the space where analysis and intuition, qualitative and qualitative, logic and emotion, overlap. In my version, data and storytelling work in concert with each other. Both modes of thinking can be incorporated into every step of the process, from determining which questions to ask, to figuring out what data and methods can answer those questions, to designing and communicating the insights.

A couple of weeks ago, I had the opportunity to spend a day in Athens, Georgia with folks — designers, journalists, technologists, teachers, researchers — who share a love of data storytelling at Tapestry, a conference hosted by visualization software company Tableau. Here are some five homework assignments my favorite speakers inspired me to give myself, along with a few insights:

Homework Assignment #1: Create a data graphic on your cell phone

Hannah Fairfield, designer at the New York Times, spoke about how to create data graphics that build to a reveal — just like any other good story does. This concept was echoed throughout the day: When you show your audience “2 + 2” and trust them to infer “4,” they experience the information as discovery, which is stickier than simply presenting an interesting finding. As part of this talk, Fairfield shared an example of on motorcycle helmets and fatalities presented as a slideshow that steps the audience through a data story toward an insight. The slideshow was designed to be viewed on a cell phone (as is every graphic Fairfield develops for the Times):

This data story from the New York Times is, effectively, sequential data art. It’s designed with a mobile-first mindset.

While this graphic was designed for a cell phone, it wasn’t designed on a cell phone. Why should you even try such a thing? Designers often don’t “think in mobile” because they do their work on giant screens. Forcing yourself to not only view the graphics you make on a small screen, but build it there as well, will help reinforce the mobile-first mindset.

If, like me, you are at a loss in terms of how to actually create and edit graphics on your phone, here are some apps to get you started: iOS, Android.

Homework Assignment #2: Use a single set of data to tell 7 (or more) types of data stories

Ben Jones of Tableau Public — inspired by Kurt Vonnegut and Edward Segel and Jeffrey Heer’s must-read paper on visual narratives — presented seven common data story patterns: change over time, drill down, zoom out, contrast, intersections, factors, and outliers.

To illustrate these story types, Jones used a single data set on the freedom of the press. This approach highlights something I truly believe is at the heart of data storytelling (or any kind of storytelling, really): A single data set can lead to infinite stories. This is far more exciting than simply telling the story through different channels or mediums: The insight Jones pulled from the data, as well as the presentation, was different in each example he presented.

There are no doubt more than seven types of data stories, but this is a great start. Images are from Ben Jones’ presentation, (which I then converted to an animated gif).

It’s easy to become accustomed to telling certain types of stories with data. Those habits help us do our work better, especially when operating on a deadline. But sometimes we need to break them. Creativity is a never-ending dance of finding, learning, and outgrowing patterns, and this homework assignment can help you navigate that process.

Homework Assignment #3: Collaborate with a researcher and offer to visualize, for a new audience, the data they’ve compiled

Working with a small team, when Popular Science’s Kathryn Peek wants to do an ambitious storytelling project, she looks outside her organization for opportunities to collaborate. Often, this means teaming up with scientists and researchers to present their data in new ways for new audiences. As they work to present the information in new ways, they often end up uncovering new insights as well, as they did with this visualization of the lifecycle of scientific ideas.

Data storytelling happens in all kinds of settings — journalism, science, business, academia, education — and each of those communities has developed its own best practices, conventions and points of view. Cross-pollinating your data storytelling with another discipline provide news inspiration and has the potential to expand your skills as a data storyteller.

Homework Assignment #4: Write down your definitions of “data,” “story,” and “data story”

Data designer Kim Rees of Periscopic introduced herself with a statement she knew would be provocative at a data storytelling conference: She doesn’t think data and story belong together.

As she spoke, it became clear that her definition and my definition of “story” didn’t line up. She equates story with fiction, whereas I see story as a way of organizing information. (Many of my favorite storytellers — Joan Didion, John McPhee, Ira Glass, Jad Abumrad — deal in nonfiction.)

The reason I bring this up is not to debate Rees — although that would be fun — but to encourage as a thoughtful approach to the work of data storytelling. Definitions needn’t be concrete — in fact, I’m sure mine will be mutable — but forcing ourselves to think about them is a great way to examine the principles and assumptions that underlie the work we do as data storytellers.

Homework Assignment #5: Tell a data story using a QUESTION → ANSWER → NEW QUESTION structure

Storytelling is an an evolutionary strategy, says Newman University English professor Michael Austin. We tell stories because they help us survive; the entertainment value is a by-product. (Austin has written an entire book on the topic, Useful Fictions.)

The best stories create, then mediate, anxiety. How? By posing questions, then answering them. If you can do this in a way where the answer of one question gives rise to a new question, well then you’ve got yourself an infinite storytelling loop.

For a great story, pose a question that demands to be answered. For a never ending story, keep doing this.

How can this help you be a better data storyteller? You know why your data is important, but your audience doesn’t. Before you present the data, present a question that can be answered by the data. Create a need and then satisfy it.

What comes next?

As I complete these homework assignments, I’ll post the results and share them here — and I encourage you to do the same.

What data storytelling homework assignments have you given yourself, and how have they helped you be a better storyteller?

Jordan Wirfs-Brock is a data journalist with Inside Energy.