How coronavirus charts can mislead us

How to read a popular chart of coronavirus cases by country.

Support Vox by joining the Video Lab at http://vox.com/join or making a one-time contribution: http://vox.com/contribute

Much of the data about the coronavirus epidemic and covid-19 is flawed. It is collected and reported in different ways by different countries, and almost certainly undercounts the number of cases and deaths. But organizations and journalists still need to report the available data to inform the public and help guide policymakers. Much of that data ends up in visualizations, like charts and maps, which can make it easier to understand and analyze.

But it’s important to know how the process of data visualization can shape our perception of the crisis. In this video, we deconstruct one particularly popular chart of covid-19 cases around the world which uses a logarithmic scale, and explain how to avoid being misled by it.

Sources:
https://ourworldindata.org/coronavirus#all-charts-preview
http://91-divoc.com/pages/covid-visualization/
https://twitter.com/jburnmurdoch
https://medium.com/@mettlinger/a-different-covid-19-graph-updated-and-revised-40f04b90c4e3
https://www.vox.com/policy-and-politics/2020/3/13/21178289/confirmed-coronavirus-cases-us-countries-italy-iran-singapore-hong-kong
https://coronavirus.jhu.edu/map.html

Read more about the Covid-19 coronavirus pandemic at http://vox.com/coronavirus

How social distancing and “flattening the curve” works: http://bit.ly/3aOlHM8
12 things everyone needs to know about the coronavirus pandemic: https://bit.ly/2xYrW2x
The differences in countries’ coronavirus death rates, explained: https://bit.ly/3bLGMaV

How has the coronavirus pandemic impacted your life? Share to help Vox’s reporting: http://bit.ly/2vBunqA

Vox.com is a news website that helps you cut through the noise and understand what’s really driving the events in the headlines. Check out http://www.vox.com.

Watch our full video catalog: http://goo.gl/IZONyE
Follow Vox on Facebook: http://goo.gl/U2g06o
Or Twitter: http://goo.gl/XFrZ5H

  • Likes: 53830
  • Dislikes: 2380
  • Tags: coronavirus,Covid-19,data viz,data visualization,chart,Vox.com,vox,explain,explainer,coronavirus vox,covid,covid-19 vox,covid19,how to read a chart,logarithmic,coronavirus chart,covid 19 chart,coronavirus cases,coronavirus deaths,coronavirus us,coronavirus explained,coronavirus graph,how to read a graph,logarithmic scale

Transcript

You may have seen this chart since the start
of the coronavirus pandemic.In one image, it appears to capture the state of each nation’s battlein the global war against the virus.But like all data visualizations, its design 
tends to emphasize some things and hides others.So here are 4 things we need to know to understand
this chart.First, this is not a chart of all coronavirus
cases.It’s only showing us confirmed cases.That means each line doesn’t just reflect
the state of the outbreak in a country, butalso how aggressively that country is testing
people for the virus.Take a look at Japan and South Korea earlier
in the pandemic.Japan’s outbreak looked pretty small in
comparison.But the available data on testing shows us
that South Korea had tested vastly more peoplethan Japan did, even though their population
is less than half as big.And now, as Japan slowly increases their testing,
the outbreak there looks more worrying.It’s a good reminder that we can’t understand
case datawithout some sense of the testing levels.And that’s especially true for lower-income
countrieswhere we know their testing capacity is limited.The second thing to know is that the scale for the y-axis on this chartis a bit different from most charts.It’s called a logarithmic or log scale.On a typical linear scale, you divide the
space by adding the same value over and over.The log scale is made by multiplying a value,
in this case, ten.100 times 10 is 1000, times 10 is 10,000,
times 10 is 100,000.That means that there’s no fixed amount
of space on this chartfor a certain number of cases.So the first 100,000 cases take up this much
spaceand then the next 100,000 cases get just this much.The higher the numbers, the more visibly squished
they become on a log scale.So why do it this way?Well, let’s take the 5 countries with the
largest outbreaks right now, and rewind themback to March 17th.On a linear scale, it looked like things were
pretty bad in Italybut the others were doing better.The log scale offered a much clearer warning:we were all on the same path of exponential growth.It’s the nature of infectious disease that
numbers get big fast.So it makes sense for numbers to get big fast
on the chart too.Fast forward a few weeks and the linear scale shows cases climbing and climbingwhile the log chart shows curves that are flattening.As governments have implemented lockdowns
and social distancing, the virus is spreadingat a slower rate than before, which isn’t
very visible on the linear scale.But keep in mind that the difference between
this dot and this dotis more than 32,000 people.And the log chart tends to downplay just how
many more confirmed cases there are in theUS than in the other countries.Which brings us to the third thing to knowabout our chart: it doesn’t account for
population size.When you adjust for population, really small
countries like Iceland and Luxembourg  appearto have the biggest outbreaks for their size,
which may reflect higher testing rates.The US and China have much bigger populations
so their curves drop a bit.But the size of a country doesn’t really
affect the growth rate of its cases, and itdoesn’t tell us much about how much the
country is struggling.It just pushes smaller countries up on the
chart and tends to hide the fact that theoutbreak is especially bad in certain regions
of bigger countries,like the state of New York.And the last thing to know about our chart
is that the x-axis doesn’t plot time bythe date, 
but by the number of days since the countryrecorded 100 confirmed cases.For Italy, that was February 24.For Turkey, March 19.When they’re all layered on top of each
other, it allows us to compare the trajectoryof the outbreaks,
but it tends to obscure the fact that thepandemic hit some countries before it hit
others.The world watched as tens of thousands of
cases appeared in China.Then big outbreaks in South Korea, Italy,
and Iran, sent a message about what was to come.Two weeks after South Korea reported its 100th
case, the United States’ did the same.In a situation where actions taken early can
have a much bigger impact than actions takenlater, time is a crucial factor. and we have
to remember that some governments had moretime than others.

Plaats een reactie