Many of our workshop participants mistakenly believe that there is really only one way to tell stories with data. Like “data stories” is somehow a genre of its own. As if the mere inclusion of statistics precludes you from choosing a specific style of storytelling.
But data can be applied to support many different types of stories. And choosing the right one is critical to crafting a data story that will attract attention and keep your audience engaged. So in this post, we’ll review seven different ways to tell stories with data.
No. 1: Change Over Time
Often, we tell stories with data when we want to convey the way something is changing over time. There are no specific constraints on the amount of time or the type of change you can include in this type of story.
This type of data story is best if your dataset covers a significant period of time and includes:
- Change that is dramatic and unusual (indicating an unexpected change), or
- More gradual but meaningful (demonstrating a changing trend)
Stories with data that show change over time are easily visualized and generally fairly easy for audiences to understand.
Example: The New York Times ran this article earlier this year about how the average American is getting more sleep today than in 2003.
No. 2: Drilling Down
For more complex trends revealed in data that you’ve spent more time analyzing, you might want to use this method for telling stories with your data. This strategy is ideal if you:
- Have multi-layered data, and
- Want to convey a more in-depth understanding of a trend you’ve observed
Telling stories with data in this way does generally require more work — you’ll need layered data and greater analysis. But it often provides a more unique (and often, compelling) story that some of the more simple narrative types.
Example: In late 2017, the New York Times published this story about the impact free trade agreements have had on Mexican health and obesity rates.
No. 3: Zooming Out
This strategy for crafting stories with data is the exact opposite of the “drilling down” approach. By shifting the focus of your narrative to give your audience a big-picture view of the issue you’re talking about, you can at significance to data that might otherwise seem meaningless.
To zoom out in your data story, you’ll likely need to incorporate broader datasets so you can shift the audience’s perspective. This type of story can inspire unification or division, depending on how you leverage its power.
Example: This New York Times story is focused on coral bleaching and paints a picture that links individual bleaching events to ongoing global warming.
No. 4: Contrast
If you’ve got a dataset that includes a wide range of variables — or more than one dataset on similar or related topics — you might want to try highlighting contrasts in your stories with data. Be warned: in many applications, this style allows your audience to apply their own interpretation to your story (which can be a blessing or a curse).
This method is one of the less rigid ways to tell stories with data. It is particularly effective if your story draws a conclusion that readers might find counterintuitive.
Example: This New York Times piece examines whether or not there is a causal relationship between bad behaviour in video games and in real life.
No. 5: Intersections
If your data story aims to add nuance to concepts or situations that your audience already has a decent grasp of, you may want to consider this strategy. It allows you to convey how different aspects of a story work together within a trend or situation.
Writing stories with data in this style is not for the faint of heart. Intersectional stories require data at a granular level and some complex analysis. This is, however, the most inclusive of all our data story types and lets you illustrate the differences between subgroups in your data.
Example: This New York Times article digs deep into the various risk factors associated with heart attacks and strokes to look at why many victims have little or no warning they’re in danger.
No. 6: Factors
Many of the most popular stories with data relate to causality — that is, examining the factors that trigger a specific event or phenomenon. This type of story attracts readers easily, but unfortunately, is extremely prone to statistical traps.
If you have a lot of data and are confident in your analysis (read: you’ve spent a lot of time analyzing your data), you can tell a powerful and compelling data story by looking at what factors influence current events or trends.
Example: In this story, New York Times writers look at increasing levels of microplastics in the world’s water sources and where they are coming from.
Example: This piece examines the health benefits of fibre and why experts are recommending its inclusion in a healthy diet.
No. 7: Outliers
If your data contains significant outliers, don’t rush to reject them. In some cases, outliers represent an error in your data collection process. But there are occasions when outliers can form the basis for some interesting new stories with data.
Data stories based on outliers are some of the easiest to write and understand. The mere existence of an outlier in your data can be interesting. You don’t necessarily need to explain why it’s there. Outlier stories require minimal data and are easily visualized.
Example: This New York Times story covers the surprising number of fossils unearthed in a single location — it doesn’t explain why so many fossils were in one place, just highlights the relative rarity of the discovery.
How Will You Tell Stories with Data?
Which of these story types works best for your data story? If you still have questions or want expert assistance telling stories with data, talk to the team at Datassist. We work with journalists, governments, and social sector organizations of all shapes and sizes. Drop us a line today.