In this week’s class reading, Lev Manovich gives us more tips about design and media. Fun. I wish Manovich wrote for the simple reader like myself. I want to understand what he’s saying but he’s writing for the “techies” and not someone like me. Sad. In a section of his book called “Representation versus Control” Manovich compares the difference between human-computer interface (HCI) and cultural interfaces. I may have misinterpreted what Manovich is saying, but he suggests that we are at a point of technological transformation where computer programmers are behind in trying to develop more efficient methods of HCI.
I wish Manovich would give an example of what he means. Perhaps he himself cannot think beyond his philosophy but feels, at the bottom of his techie heart, that there is a better way of creating cultural interfaces that would help to represent our memories, values, and experiences.
It may be pointless to debate Manovich’s arguments since he wrote his book in 2001. A lot has changed. I went from using CD players and an IBM desktop to surrounding myself with everything Apple (and a crappy Motorola touch phone that doesn’t compare to my beloved iPhone but does the trick until I figure out how to unlock my silly iPhone4. Suggestions?)
But if I took to heart Manovich’s points on design and applied them to data journalism, then I think he still has a relevant argument. Data journalism is still “new” to our profession. It’s starting to spread beyond the techie journalists, but the way that we look at data may be the same approach as it was three or four years, which is considerably old in journalism at a time when news corporations are quickly adapting to use of smartphones, tablets, etc.
Data journalism should be designed in a way that takes advantage of the new technology. But it’s not. Most visualizations and graphics are 2-D with little interactivity. Yet this is the Age of Big Data, according to the New York Times, so we should reflect this change in our designs because readers will expect more than a simple graphic.
This is exactly why I want to become a data journalist who is (slowly) becoming proficient in computer programming. Analyzing data is hard work and the payoff is small if people don’t read it. So golden rule #2 in data journalism design: people like pretty, shiny things. For example, readers who are interested in the primaries may like this design of Maine’s caucus results by WNYC, but they’ll love what Slate created, and will probably spend more time reviewing it.
Shiny, pretty data visualizations are fun and will likely attract more readers, but journos shouldn’t get too preoccupied with including all the neatest tricks so that the visualization becomes too complicated to understand. This is the view of Donald Norman in his book, “The Design of Everyday Things”.
It’s too bad that the creators of Bear71, the data project I reviewed last week, didn’t take note of Norman’s view. Bear71 exemplifies what goes wrong when a data journo is too caught up with her design, forgetting who her audience is.
The audience, Norman seems to suggest, is an integral element of a design that many designers often overlook. While it’s difficult to plan for everyone since there’s no such thing as an average person, the design of a thing (in this case, a visualization or a graphic) should try to be inclusionary rather than exclusionary. And that’s why Slate‘s fun little interactive design of the Republican primaries works. It’s simple, it gives readers a clear understanding of who the political candidates are, how they compared to each other, and even shows a small pop out of each candidate’s face to avoid confusion.
And the animated visualization continues to be useful because it’s designed to be “live” through regular updates. If this was a 2-D graphic, I would yawn and change the webpage because in three weeks (the published date is January 23 2012), candidate Rick Santorum has pulled ahead of Mitt Romney. People will continue to view this design because it’s fresh.
It’s also fun. The design is a literal translation of the Republican primaries as a horserace. The designer, Will Oremus, knows who Slate readers are: politically engaged and technologically savvy. At the same time, Oremus made it simple enough to attract the attention of readers who may not follow the day-to-day developments of the primaries but still know enough to not have to be told who each candidate is. The design of this visualization may not have worked in The Globe and Mail, a Canadian daily, but it works well in Slate.
Data journos, therefore, must straddle between the simplicity of a design while maintaining its technological relevance. Simplicity is key. Shiny things also help. The end. Is there really anything more that data journalists should consider when thinking about the appearance and usability of their product? Probably but I’m sticking with these two golden rules for now.