Archie Tse's over-sized laugh fills the room. He is busy defending the little-used, half-moon chart as the best way to visualize the recent Congressional elections and maybe the answer to any graphic challenge you can imagine.
Tse (the graphics editor for
The New York Times, a Poynter visiting faculty and National Advisory Board member),
Laura Stanton (The Washington Post)
Juan Thomassie (
USA Today)
are leading an
informational graphics seminar through a post-mortem of mid-term election graphics. A conversation that began with a discussion of the merits of color-coded geographic national maps versus a cartogram has gotten more technical quickly.
Telling a complicated election story requires more than debating the efficiency of a half-moon graphic. For Tse and Stanton, the challenge is to sift through the numbers and trends to show something that is both rich and deep, but also clear and easy to understand.
Most specifically they are dissecting a full-page election graphic
The Times produced to show the shift of voting patterns in the House races from Republican to Democrat.
Tse outlined his process starting with a series of visualizations of House results. In the weeks before the election, members of
The Times graphics staff tested different forms of the graphic, plugging in election data from as far back as 1968. They worked forward testing each result.
Some of the mockups, including these scatter plots
are fairly complete but each one was ultimately rejected for various flaws. Each graphic form tried to show the transition of voters from Republican candidates to Democrats. Tse was convinced that there was a way to show the movement of individual districts at the same time demonstrating a national trend.
Finally, Matthew Ericson of
The Times tried applying a form he saw in a scientific study. Tse realized they might finally have found the way to show the trend. At first glance, the multiple layers of information and the unusual presentation could make the resulting graphic daunting. Understanding the need for simplicity and clarity, Tse, Ericson and their
Times' graphics colleague, Amanda Cox, spent extra time creating the key, explainer text and breakout information. The goal: help readers comprehend the chart.
The chart itself shows each district represented by an arrow. The arrow points in the direction of the change and the length of the line represents how far the district has changed. In analyzing the graphic the group raised some concerns: the arrows are fairly small and the lines are light, which can make reading difficult.
The core of the graphic is the vertical, colored chart. The chart is divided into seven horizontal bands, showing the strength of seats for each party. At the top are very strong seats -- those facing minimal opposition -- for Democrats. As the chart leads down the page the strength of Democrats seats lessens. In the center are the seats still undecided at the time the graphic was published. At the very bottom of the chart are the very strong Republican seats. These delineations are made clear by white dividing lines. Each section is described in text and an oversized numeral telling how many seats are in each category.
The 28 seats that switched to the Democrats are called out in a thicker blue line and identified by district number. This emphasizes the key races and offers more information about them.
Although there was near-universal admiration among the seminar group for the work, there was some concern that the information was too dense and required too much effort to read. Tse's editors expected a graphic simple enough for a high school student to understand.
He compared the election graphic to a 3,000-word story. Perhaps only those readers with intense interest in the topic would devote the time and attention required to read the story. Just as newsrooms should not shrink from providing that kind of depth in words, neither should they hesitate to tackle complex topics with graphics, Tse argued. The trick is to render the journalism in both forms -- graphics as well as words -- as comprehensible as possible.
Feedback: The New York Times
’ graphic showing the Democratic shift was one way to tell this story. Do you think it worked? What, if anything, could have been done to make it better? Or did you see, or create, other election graphics to tell a similar story? Send me examples of the work you did or saw and I will add them to this page.
