Data Visualization: Story Telling

As the Stanford paper says, "data stories appear to be most effective "when they have constrained interaction "at various checkpoints within a narrative, "allowing the user to explore the data "without veering too far from the intended narrative." In other words, they're not exploration tools, but rather, narrative experiences that provide context and direction, not just a pile of numbers and charts.

Contents of a Great-Story

The beginning sets the context for what we're about to hear, the middle explains what's going on, allowing us to wander and explore the universe of ideas being presented and get introduced to any anxieties that need resolving as part of the story, and the conclusion, hopefully, resolves those anxieties, teaching us how to solve problems and what transformation characters have to go through to do so. This logical flow is almost the only form a great story can take.

Finding the story in Data

The thing you want to focus on is thinking in terms of story while you're analyzing your data. The questions you're asking and the answers you're finding, will need to be able to be assembled into a logical flow with a focused storyline. So think about that flow while you're even doing your analysis. One discover interesting things in data and one shouldn't be afraid to bookmark findings that may not fit the flow. Even if the data is interesting. Perhaps one might have two stories to tell. And that's more than okay.

Outline your story

The idea is to figure out what comes first, what's in that piece of the story, and then what happens? And then what after that? The purpose of this outline is to help figure out how long the story is, what the key topics are, and whether or not they flow logically from one to another.

Sketching the content

· Vet and test the ideas · Confirm your story structure and data points that will communicate your data in the best way possible · Failing early and adjusting and failing again until one hits on something that works. · Understanding and thinking about what visualization will work out the best

Flow-Diagram

Flow diagrams have been used for a very long time to visualize things like org charts, database diagrams, and business processes. And the interactive version is really a interesting recreation of it, because the interactivity allows one to find the stories more easily. If the data describes a flow from one thing to the next, or some sort of hierarchy, or if you can use the metaphor of a flowing process, or if you're telling a story of proportionality, especially successive proportionality, using a flow diagram might be a great way to tell your data story.

Personalization

Everyone is somewhat of a narcissist because everyone's interested in things that are about themselves and the things that they care about. The story in its consumption is directly about the listener and their reality as much as it is about the storyteller. If you can take the data that you're communicating to your audience and transform it into a personal experience for them even if it's not their personal data, you'll allow them to see themselves in your data and you can have a very similar effect.

Progressive Depth

Story telling with four-by-four model presents content to people in levels. Let’s call them water cooler, the cafe, the research library, and the lab. The water cooler moment is essentially that initial attention-grabbing piece of content. It's the thing that lets people decide whether or not they're interested in what you're talking about. Just like if you're chatting at the water cooler, people around you will decide if they want to hang out and talk longer. If they do, they might go to the cafe with you and have a cup of coffee. The water cooler moment is like an image or a headline. Cafe content is like the blog post or the short article or a three to five-minute video. If they're really still interested in the topic after consuming that, they'll go to the research library. They will dig deep, they'll read the a hundred-page PDF. And finally, you have the lab experience, which is like that interactive data experience, the dashboard where people can explore the content much more deeply and maybe on their own. Those are the four levels of content. The point is that if you create longer form content, you always have to create water cooler moments to attract attention that lead to cafe content, to build context and knowledge, followed by whatever depth is still available after that. It's all about stepping people into content and letting them self-select based on their interest in what you're talking about.

Stories make an emotional connection in a way a collection of facts really can never do and data stories are extra compelling as they mix the evolutionary imperative of stories, with the important factor of an element that we need to make great decisions.