BEWARE OF THE INFOGRAPHIC


WPCNR BOOKMARKS. By Savannah Jacobson, from the Columbia
School of Journalism. Oct. 29, 2019:

Alberto Cairo is on a mission to improve how journalists use charts.

“Visualizations, charts can be incredibly powerful at exploring data,” he
told me recently. They can also be powerful as tools for communicating
information to news readers. “If you know how to use them well,” Cairo added. To his endless frustration, too many reporters do not.

In his new book, How Charts Lie: Getting Smarter about Visual
Information, Cairo, who is the Knight Chair in visual journalism at the
University of Miami, aims to dispel the myth of objectivity, and the air of
truthfulness, that has been undeservedly awarded to numbers.

A chart, he said, is a “visual argument” that is only as strong as the data
on which it’s based. To tell a reliable story with a chart requires an
understanding of its data—what it consists of, how it was gathered, who it might leave out.

“We journalists are mediators,” Cairo explained. “Mediators between
science and complexity, and the general public.”

Throughout the book, Cairo breaks down common mistakes journalists make.

First up: assuming that correlation indicates causation. To demonstrate
why that’s wrong, Cairo produces a chart, using data from the World
Health Organization and the United Nations, showing that cigarette
consumption by country is positively correlated with life expectancy.

“I have seen graphics like that described by journalists—including myself because most of these things are mistakes that I have made myself—
describing this kind of chart as ‘the more we smoke, the longer we live,’” he told me. But in reality, he writes, “a chart only shows what it shows, and nothing else” (emphasis his).

Cairo then breaks the data down further, with fifteen more charts,
grouping countries by income and region. Ultimately, he shows, the
original chart cannot prove anything definitive—it can merely point to a pattern.

His step-by-step instructions, at risk of becoming dry, are livened up with humor (a data point about the glam rock band Poison has no place in a
chart about heavy metal, he argues).

“I try to do it in a way that could be used as a template by translators,
communicators, journalists, to do the same thing,” he told me.

When journalists are wrong, Cairo warns, there can be serious
repercussions. A town in danger of storm damage, for instance, may fail
to take proper precautions because a broadcaster misinterpreted a
graph.

Just look at Hurricane Dorian projections: the National Oceanic and
Atmospheric Administration relies on a graphic—called the cone of
uncertainty—to explain potential paths for major storms and, during
Dorian coverage, many reporters interpreted the cone graphic as
showing the entirety of the storm’s wrath.

That left out a lot of possibilities. “I have seen TV newscasters explaining
this map wrong and it drives me crazy,” Cairo said. “Like saying,
‘Oh you’re outside of the cone, you may not be in danger.’ Well that’s
actually not true.”

Cairo doesn’t want to put journalists off charts, and he has ideas about
how to produce them effectively.

When news outlets design their own graphics, Cairo suggests, they should introduce a “me-layer” into the design. Why was the New York Times
 dialect map so popular? “Because people see themselves in the data,” he writes. “And they see their families in the data, and they can compare the way they talk with how other people talk.”

Perhaps most important, Cairo writes, reporters shouldn’t assume that
visuals serve as a substitute for words. Sometimes, a lengthy explanation is what’s required. At the same time, when reporters are trying to make a point, they need to just spit it out:

“If you really want to emphasize something, emphasize it,” he said. “So
people will not miss it. If there’s a particular pattern, or a particular data
point or a particular fact that should not be missed, just show it.”

And when all else fails, ask a data scientist. “You need to basically give
them whatever it is that you’re writing,” Cairo advised, “and very openly say, ‘please destroy it.’” 

Below, more on Cairo and how journalists use visual data:A chart with dubious political categorizations of media outlets, and reportedly being
taught in media literacy classes, spread around the internet this week.

Good news: you can ignore it.

“The main reason this chart is so deceptive,” Cairo writes on his blog, “is
that it compares things that aren’t comparable. Come on, Breitbart or The Federalist rags at the same level of ‘bias’ as Vox? The Washington
Examiner at the same level as NPR? Those aren’t equal. Neither in terms
of trustworthiness, nor in terms of ideological bias.”

“His book reminds readers not to infer too much from a chart, especially when it shows them what they already wanted to see,” The Economist
writes in a review of Cairo’s book, noting that he has sent a copy to the
White House. 

ICYMI: CJR hosted a series of Q&As with visual journalists who work with data, including Quartz‘s David Yanofsky and ProPublica‘s Lena Groeger.

Comments are closed.