A data visualization expert answers 5 key questions on coronavirus graphics

The spread of COVID-19 has brought numbers, statistics and charts to the forefront of the daily lives of millions of people. Whether they’re tracking infections for public health responsibilities or simply taking a detailed interest in one of the biggest stories for more than a generation, there is a huge demand for reliable data.

John Burn-Murdoch, a senior data-visualization journalist at the Financial Times, recently took to YouTube to explain his approach when compiling charts on the coronavirus outbreak.

1. Why use the logarithmic scale?

“Viruses don’t spread exponentially,” says Burn-Murdoch. “They don’t go from one person infected today, then two, then three, then four. It's more like one, then two, then four, then eight – rising at an ever-increasing, ever-accelerating rate.”

Whereas a linear scale graph will produce a near-vertical line to represent the exponential spread of the coronavirus, logarithmic scales even out the rapid growth making it easier to interpret the data.

“We want to inform people and make them aware of the severity of the issue but not to panic them. By showing this on a straight line, we're emphasizing that there's an inevitability about how coronavirus spreads,” Burn-Murdoch explains.

2. Does adjusting for a country's population size make the data easier to read?

In its representations of the spread of the coronavirus, the Financial Times has not taken population size into account. “This,” says Burn-Murdoch, “is more of a judgment call.”

There is data available that shows the spread of coronavirus per capita, or per million people. But that can produce some potentially misleading results. A very small country with just a few infections will rank far higher than a larger country that might be struggling to contain the outbreak. The overall population of a country does not limit how fast the virus spreads. That is determined by the way people in that country interact.

Burn-Murdoch says the Financial Times is focusing on trajectory. “We're focusing on saying, ‘Where are things right now? Where are they going to be in a few days? And how does this compare to other countries that you're already familiar with from following the news?’”

3. Why is it important to talk about the number of confirmed cases?

The number of confirmed cases of COVID-19 in any given country will depend on several factors. One of which, inevitably, is the amount of testing that country is doing.

South Korea, for example, has tested in excess of 350,000 people. And last month, the Director-General of the World Health Organization, Tedros Adhanom Ghebreyesus, said: “We have a simple message for all countries: test, test, test.”

Not every country has embraced his call – not every country has the resources to carry out mass testing. All of which means the number of confirmed cases in one country is not directly comparable with that of another.

“There are going to be hundreds, thousands of people in countries all over the world who do have coronavirus,” says Burn-Murdoch. “They may be completely symptom-free, but they've not yet been tested. And so, in the earlier versions of our charts, showing the case trajectories, the Y-axis title talks about the cumulative number of cases.”

4. Why show the lockdown date on a chart?

Somewhere in the region of 3 billion people are living under lockdown conditions, or have been until very recently. The Financial Times’ charts include the date each country implemented these kinds of restrictions.

Tracking the effectiveness of lockdown periods on the spread of the virus will be an important indication of how well public health responses around the world worked. But Burn-Murdoch sounds an important note of caution.

“The number of deaths in a country isn't going to start flattening overnight... because it takes two weeks or more for someone to go from being infected with the virus to dying, if they, unfortunately, reach that stage.”

5. What is the data telling us about the nature of the spread?

Viruses tend to spread in small clusters that gradually grow, leading to more clusters. That was seen in the earliest clusters in Wuhan and nearby cities in Hubei province. Latterly, New York has become a hotbed of infection.

Looking at how different regions are affected, rather than countries as a whole, can tell that story, as Burn-Murdoch explains. “When you think about the impacts of the lockdowns that are now being put in place, we're really talking about cities, which are usually these vibrant, dynamic, busy, noisy hubs falling silent.”