A data designer’s responsibility during a global crisis

If you’re like many of us, you’ve been obsessively refreshing the news and Coronavirus maps for at least a week now. Or, if you’re like me, you’re making an intentional effort to control your media intake during this anxious, confusing time.

While most of us are on the receiving end of the news, simply waiting to know what happens next, some designers are working carefully to disseminate it.

Presenting information in a clear, honest way is important right now. It's a fine line to walk between relaying the gravity of the situation and not contributing to more panic and chaos. The news cycle is moving faster than ever to cover the quickly spreading virus, providing an overwhelming amount of (often conflicting) information to sift through. Publications like The New York Times and The New Yorker, which typically put content behind a paywall, have provided free access to all Coronavirus updates.

Designers have a responsibility to visualize information in a simple, accurate and easy-to-understand way. This couldn’t be more true for designers reporting the news during this global crisis.

A simulation from the Washington Post, which illustrates the exponential effects of a disease and how it spreads, has been making its rounds over the last couple weeks. It’s been a challenge convincing people to stay inside their homes, and this visual makes it clear how one person’s actions can affect countless others. It's currently the most-viewed article – ever – on the Washington Post website.

Washington Post graphics reporter, Harry Stevens, who created the visuals, originally planned to simulate Covid-19 itself. After speaking with a professional disease modeler, he realized it would need to be more simple. The resulting simulation, which shows a sequence of moving dots that bounce about the screen, is simple yet powerful.

“It is hard to strike the right balance, but for this piece I feel we hit the nail on the head,” said Stevens. “I have gotten hundreds of emails over the last few days from doctors and mathematicians, as well as everyday folks expressing gratitude for the graphics.”

Amanda Makulec, data visualization lead for the technology consulting firm, Excella, said Stevens’ graphic may be one of the best data visualizations and pieces of public health behavior change messaging she’s seen.

“The day it was published, I had at least a dozen people send me the link and many more posted it to social media,” Makulec said. “As a result, a key public health message (stay home!) was amplified quite effectively.”

When one infographic can influence public behavior and potentially saves lives, the weight of a designer’s responsibility is greater. Considering the volume of information, both now and before the pandemic, it’s easy to scan headlines or graphics and make incorrect assumptions. It’s also easy, whether intentionally or unintentionally, to mislead readers through design.

Designers are taught to present information in its simplest and most aesthetically pleasing form. When it comes to data, those inclinations and good intentions can lead to mistakes.

Design can lie

In Alberto Cairo’s book, “How Charts Lie: Getting Smarter about Visual Information,” he explains how graphs can be skewed and misinterpreted.

“Charts lie in a variety of ways — displaying incomplete or inaccurate data, suggesting misleading patterns, and concealing uncertainty — or are frequently misunderstood,” Cairo writes. “To make matters worse, many of us are ill-equipped to interpret the visuals that politicians, journalists, advertisers, and even our employers present each day, enabling bad actors to easily manipulate them to promote their own agendas.”

Everything from a designer’s choice of chart, scale, color and even typeface selection affects how it’s perceived by a reader. A chart zoomed in too far or shown on the wrong scale can make a trend seem more alarming than it truly is. The color red invokes fear or alarm and can be used to manipulate a response in readers.

Designers are inclined to simplify. We are taught to present information in its simplest and most aesthetically pleasing form. When it comes to data, those inclinations and good intentions can lead to mistakes.

“The biggest mistake I see in simplifying graphics to suit easy social sharing is to under represent the uncertainty in the data,” Makulec said. “We remove error bars and ranges in favor of straight bar charts, or plot a line of numbers that really should have a band around it, or bury notes and issues with a low response rate to a survey in the footnotes where readers seldom notice.”

The fact that people scan, rather than read, poses a dilemma for a data designer: How does one make a graphic easily and quickly digestible at a glance, while not disregarding or concealing necessary information?

On COVID-19 charts, for example, “confirmed cases” may take more space than “cases," but that modifier matters. It’s a balance between staying brief and simple without losing meaning.

A detailed chart is not necessarily a factual chart, and the most beautiful, easy-to-understand chart can still be misleading.

So how does a designer create a responsible visualization?

Naturally, it requires an understanding of the data.

Stevens compares his process to any form of reporting, the difference being that with graphics reporting, one spends more time interrogating the data than interviewing sources.

When creating his simulation for the Washington Post, he first interviewed sources and read published papers on the subject. Working off an experiment he'd done previously with collision detection, he created his dots simulation. With the help of his editors, he then refined his working prototype over a week or two before it was published.

The goal and audience, as with any design, also determines how a graphics designer approaches their visual.

“I always start with who the audience is and what I want to communicate with the visualization,” said Makulec. “Balancing who will read the visual and what I want them to take away from the chart is what leads my decision process.”

From there, she sketches rough pictures or creates quick prototypes.

As both Stevens and Makulec explain, feedback is an invaluable part of the process. Seeking feedback from someone uninvolved in the design process, who can look at a prototype and say what their key takeaways are, is a gut-check for how well your visual communicates.

When asking for feedback, Makulec includes the project goals, audience and stage, as well as what type of feedback she’s looking for.

“If you’re not a subject matter expert in the topic you’re creating visualizations and writing about, ask for feedback from someone who is, in order to make sure you don’t lose technical nuances,” said Makulec.

It may seem like a straightforward enough process, but search "COVID-19" on Google and the amount of information and misinformation is formidable. A detailed chart is not necessarily a factual chart, and the most beautiful, easy-to-understand chart can still be misleading.

"We need to take responsibility for the ways data visualizations make information feel more certain to readers, and do our best to communicate both what we know and what we don’t in charts."

Data must be used to tell the truth

It's a reporter's job to present the truth. With the exception of opinion pieces, people expect news to be unbiased. Yet a designer’s personal experience or beliefs may consciously or subconsciously change how they present information, skewing the truth.

As Swedish physician and academic, Hans Rosling, wrote regarding the Ebola virus in his book, published two years ago, “Data was absolutely key. And because it will be key in the future too, when there is another outbreak somewhere, it is crucial to protect its credibility and the credibility of those who produce it. Data must be used to tell the truth, not to call to action, no matter how noble the intentions.”

And so a relentless loyalty to the data is necessary for a factual, informative graphic. That includes the data we know and the data we don’t know – and right now, there’s a lot we don’t know.

“We need to take responsibility for the ways data visualizations make information feel more certain to readers, and do our best to communicate both what we know and what we don’t in charts,” said Makulec. “Particularly in crises like a global pandemic where confusion can invite panic, and understating the gravity of the situation can invite complacency.”

The Washington Post piece may have inspired action, but it did so by presenting factual information, not a plea. The goal is not to make a decision or form a conclusion for a reader, but to inform or encourage further thought.

Stevens defines the difference in approaches as exploratory vs. explanatory.

“Generally speaking, if it is explanatory, you just want to help your reader understand a concept as clearly as you can. If it is exploratory, you are building an interface that encourages your reader to learn things themselves and find things you had not anticipated.”

The simulitis graphics were more explanatory than exploratory – although Stevens admits there can be overlap between the two.

"Collaboration, iteration and feedback are important parts of the design process at any time, but particularly when visualizing sensitive data in the midst of a global crisis."

Aside from the Washington Post simulation, Makulec points to several other examples of responsible and effective data visualization.

Lisa Charlotte Rost at Datawrapper created and continues to add to a list of ‘responsible’ charts around COVID-19. The page is transparent about why they made certain design choices, such as consistent use of color (intentionally avoiding the color red), answering individual questions rather than attempting to serve all needs, using clear reference lines and bands where the data is questionable, and clear headlines with supporting text for clarity.

John Burns-Murdoch, data journalist at the Financial Times, has been creating daily charts of COVID-19 case information, which Makulec considers another excellent example.

With these charts, the annotation layer tells you key information in the headline and includes annotations on the individual marks.

The chart also uses a log scale for the y-axis due to the exponential growth curve for COVID-19 (avoiding the hockey stick shape we’d see with a linear scale) and aligns the countries along the x-axis based on a common measure (here, number of days since 100th case) to better enable comparisons on a trajectory.

Perhaps most importantly, according to Makulec, Burns-Murdoch actively seeks input and feedback from subject matter experts and iterates over time.

“Rather than leaning back on own expertise, he looks for input,” Makulec said. “Collaboration, iteration and feedback are important parts of the design process at any time, but particularly when visualizing sensitive data in the midst of a global crisis.”

It’s human nature to let our fears and biases influence how we share and interpret information, both as designers and readers. The question is: How much responsibility falls on us as the reader, to think, read and repeat information carefully, and how much falls on the designer?

"As the world seems to clamor for ‘real-time’ updates on COVID-19, I think we need to pause and ask if that’s really the information we need as a general public."

Is the responsibility on the designer, the reader or both?

To Stevens, it’s the designer’s job to ensure information is understood correctly.

“I like to say that there’s no such thing as user error,” said Stevens. “If someone tells you your graphic isn’t working, it’s your responsibility to fix it.”

Makulec believes designers and readers have a shared responsibility. But for the reader, our responsibility extends beyond careful reading and consideration. It’s about being patient and thoughtful about the news we seek in the first place.

“As the world seems to clamor for ‘real-time’ updates on COVID-19, I think we need to pause and ask if that’s really the information we need as a general public,” Makulec said. “Somehow we’ve become accustomed to instant access to information, but what will functionally change about my behavior based on the number of U.S. cases at 10 a.m compared to 4 p.m.? Is this data actionable to me?”

Real-time information is actionable for other groups: hospitals and medical professionals in regions with increasing cases, local governments managing their own response to the disease, experts at the CDC and other global bodies who are actively responding to the crisis, Makulec explains. But the difference of a few hours isn’t going to change our decision to stay home or not. Waiting for information may also translate into better quality information.

Where designers can help

Everything that’s happening now is unprecedented. We know about disease but we’ve never, in our lifetime, seen it on this scale. The government is working off information as it unfolds in real-time. Scientists will be testing vaccines for the next several months. The rest of us, well, we watch from behind our windows and we wait.

Designers, however, can play a role in bringing clarity in a confusing time. But unless your job requires it, designing graphs may not be the most productive contribution.

“As designers, we need to be very cautious about designing charts and graphs with the case data, which has so much uncertainty and sensitivity,” said Makulec. “I’m enthusiastically sharing articles and information, but I don’t need to create any of my own charts when there are great ones already out there.”

Instead, Makulec recommends designers direct their enthusiasm and energy in two ways: By signing up to volunteer to work with a civil society organization that has subject matter expertise and a specific need, or creating charts that help us better understand all of the other things happening in the world during this time of social distancing.

If you do decide to create some visualizations with the open case data, read these ten considerations before you press publish.

Thank you to Amanda Makulec and Harry Stevens for taking the time to share their knowledge. To learn more about the role of a journalist during this crisis, watch Harry Stevens' "How to be a journalist" video.

Cover image created by the Washington Post

Books referenced:

Factfulness: Ten Reasons We're Wrong About The World - And Why Things Are Better Than You Think
How Charts Lie: Getting Smarter about Visual Information