Data Visualization: Effective Illustrating or Redundant Rendering?
Despite the clear benefits of visualizing information, the concept has often been treated as a last-ditch approach, a less intellectual way of arranging data, or the information equivalent of putting lipstick on a pig. That doesn’t mean we don’t want to do it, though — meme-style trends like infographics and interactive maps have caused visualization to run rampant, but in turn, have aided a self-fulfilling prophecy: the practice is mainstream now, but redundant, rendering a lot of the available visualizations ineffective.
Here’s how to determine if visualization is worth the effort — or, if you’re just getting that data all dressed up with nowhere to go.
What Difference Does it Make?
Historically, discovering great solutions begins with asking great questions. While we know that humans are primarily visual creatures, that’s not enough to justify representing information in a visual way — there should always be a return on whatever investment is put into mining and shaping data. Says Dell strategist Jim Stikeleather, “Ultimately, data visualization is about communicating an idea that will drive action,” and “for information to provide valuable insights, it must be interpretable, relevant, and novel.”
So, is it? Is the data complete, robust, and applicable to a problem? Will visualizing it provide a unique perspective that reveals new patterns and trends? Will it make the information easier to understand? Can it be clearly and wholly presented visually, or will doing so hamper the message?
If your data doesn’t require illustration to be more effective, don’t waste your time.
Old Adage, New Context
The biggest problem we’re seeing with current data visualization practices is misrepresentation. Even shaky stats can turn into an impressive, attractive illustration with the right graphics — but that doesn’t make them valuable, or even accurate.
Take political polling. Like with any data collection, polling is done using only a slice of a particular population. Myriad factors influence how participants will respond to inquiries about party preference — everything from obvious stuff like affiliations and demographics, to very marginal things, like whether they’re responding to a human being or a bot, whether the poll is via phone or internet, or even whether they had a terrible morning. Massive amounts of results get boiled down to create those simple red-versus-blue charts. They make possible election outcomes seem clear and easy to predict, despite the extensive amount of info that was discounted in order to get there.
Amanda Cox, Graphics Editor at the New York Times, cautions people to be objective about things like infographics or even simple charts, since the interpretation of data can drastically change depending on which perspective it’s filtered through. “It would be silly to interpret any data viz as truth,” she says. “They are interpretations of truth.”
In other words, don’t believe everything you see, because you probably aren’t seeing everything.
Only the Best (Practices)
If your data makes the visualization cut, there are some key things to keep in mind as you decide what type of illustrations and graphics to use and which info to include to make a greater impact. Our favorite Do’s and Don’ts:
- Don’t trick out your graphics at the expense of clearly displaying info.
- Don’t get text happy. If the info can’t be illustrated, it shouldn’t be.
- Do pull out and highlight numbers whenever possible; they speak volumes.
- Do consider physical limitations like color-blindness, and adjust your contrast accordingly.
- Don’t lie. Seriously. Being transparent and noting sources only strengthens your presentation.
- Do drive conversation by including thought-provoking questions with your graphics.
- Do consider your audience demographic, and allow it to dictate design.
- Don’t cherry-pick your data. Hiding information is just as bad as faking it.
When data visualization works, it can be one of the most powerful ways to capture and present information. When it doesn’t effort is wasted, data gets buried, and careful research is disrespected. Choose your battles wisely, and let the data design itself.