Join Doug Rose for an in-depth discussion in this video Introduce visuals, part of Data Science Teams: Telling Stories With Data.
- There are a number of great resources on data visualization. One of them is Storytelling with Data by Cole Nussbaumer Knafflic. Another is The Visual Display of Quantitative Information by Edward Tuffe. Both of these take a very tactical view of storytelling. They talk about how great visuals can create great stories. They imply that data visualization goes hand in hand with great storytelling. You'll see this throughout both the books. Storytelling with Data has six lessons.
The last lesson is Telling a Story. This comes after earlier lessons on choosing a display and eliminating chart clutter. You should think of these lessons in reverse order. You should think about your story and go back and create better data visualizations. These books are terrific, but they overstate the importance of data visualization. Charts and reports can certainly help you tell the story, but it's the narrative of the story that connects your audience to the larger meaning.
The visualization in itself, is just a small part of that effort. Your data visualizations might have just as much of a chance harming your story as helping. They can be a distraction. Every time you show a new image, it takes some time for people to process the information. They need the time to make sense of the data. That's why you need to be very conservative with how much data you display. These books give you a lot of great ideas on maximizing data while minimizing clutter.
If you use these principles, your audience will only need a moment to glance at the visualization. Then they can go back and start listening to the larger story. The most important thing to remember is that data visualization is the sauce and not the meal. A really interesting story won't need good visualizations. At the same time, even the best data visualization won't cover up a boring story. There are a few things that you can do to help simplify your visualizations. That way they can add value to your story without distracting your audience.
The first thing you need to remember is that you want to break your data into small, digestible pieces. The more time your audience focuses on the data, the less energy they have listening to your story. Try the tips in these two books to create very lightweight visualizations. Both of them describe a process of stripping away needless information. You're trying to get the bare minimum that you need to communicate something interesting. You'll also want to use high contrast colors. It shouldn't take too much effort to look at the data.
You can try turning off the text labels when you look at the visualization. See if the data still makes sense. If you choose to display a lot of data, it's better to have a consistent flow of digestible chunks. So break your data down into simple, bare minimum charts and display them throughout your storytelling session. Make a clear distinction between presenting data and storytelling. You can hold a clicker in your hand to change slides. Get the audience used to the fact that when you're holding the clicker, you're presenting data.
When you put down the clicker, you're back into storytelling. This will help your audience follow the flow of your presentation. Finally, it's very important to remember that data visualizations in themselves won't get you very far. If you want to tell a great story about a great city, like Chicago, then you wouldn't display a beautiful subway map. You'd tell stories about the great food, wonderful neighborhoods, and the brown, sandy beaches of Lake Michigan. That's the thing that gets your audience's attention.
These are the stories that make people want to visit. A wonderful data visualization, like a subway map, can tell you where to go, it just doesn't give you a reason to go there. It's a mistake to think that a beautiful visualization can spur your audience into action. If you understand the limits of visualizations, then you can certainly get some value from the added benefit. Just don't make the mistake that good visualizations can take the place of an interesting story.
- Structuring a data science story
- Defining plot, conflict, and details
- Going beyond reporting
- Knowing your audience
- Working with data
- Introducing visuals
- Eliminating distractions
- Incorporating metaphors
- Motivating the audience
- Avoiding pitfalls