Visualising COVID-19 – the challenge for data researchers
” Visuals are essentials. Sadly, there is a real human tragedy behind the statistics, the way we design and structure the narratives should account for that.”
— Yaryna Serkez, Graphics Editor at the New York Times
Beyond the visible
The streets of Sydney’s CBD are seldom so empty, with only a few people walking and jogging, shops and stores closed. The popular items people want to buy have become masks and hand sanitisers. The underlying truth is visible in the death tolls caused by the COVID-19 virus.
The World Health Organization and local governments around the world set up COVID-19 restrictions that keep people at home, unless they are essential workers, shopping for food, or getting exercise near home. Nobody knows when the global pandemic will end, but everyone wants to be prepared when it comes to protecting themselves in these uncertain times.
Interest and attention in data visualisations have surged during the worldwide spread of COVID-19. People desire to see and understand the current situations related to the virus: How case numbers are rising, which areas have more cases, what impacts the virus might have on themselves and their communities, and what they can do to protect themselves and slow down the spread.
“One of the differences between our dashboard and many of the others is that it gives you a sense of the temporality of the data in the spatial domain so you can see not only how many cases have existed in a particular location but also how long it has been since the last cases were there,” said Associate Professor Adam Dunn, head of the discipline for biomedical informatics and digital healthat the University of Sydney,who is one of the leaders in this team.
Such dashboards describe the overall situations of the spread of COVID-19 in the community. By mapping the COVID-19 data by postcodes, they highlight areas at higher risks of community transmissions. This allows the government to measure quarantine policies or make decisions about when they might be able to reopen schools, businesses, and borders, and where they should be sending more resources.
“if we see places in NSW where there have been increases in the number of cases, then we might want to consider sending more resources there. That might mean ensuring people are physically distancing. It might mean sending more protective equipment or it might mean talking to the people who run businesses in those locations,” Prof Dunn said.
Professor Adam Dunn screen share NSW COVID-19 dashboard
Visualising COVID-19 improves public transparency. People understand the risks around them in their community, which can reduce the level of anxiety, panic, and fear in the community about catching this disease, according to Prof Dunn.
“if you are in a place where there hasn’t been a case of COVID-19 for the last few weeks, then you can feel more comfortable about doing your usual activities, like going to the shops, meeting friends, because you’re less likely to be in contact with someone who has COVID-19,” he explained.
But most data visualisation, including this dashboard, is not able to visualise how often people move from place to place.
“An indication of one place being COVID-19 free for a period of time doesn’t necessarily mean there are no people from another postcode visiting those postcodes,” PROF DUNN said.
As mentioned in the attached thread, I see communicating exponential growth (virus spread) as one of the key challenges to helping people understand #COVID19.
Instead of creating visualisations of COVID-19 case data, Amanda Makulec, the Senior Data Visualisation Lead at Excellawith a strong public health background, focused on writing about how to visualise COVID-19 data responsibly, given data quality issues with the confirmed case numbers, and on helping people better understand the process of data production and interpretation.
.@COVID19Tracking launched a new set of charts disaggregating COVID-19 data by race, further underscoring the disproportionate impact on communities of colour.
Two states still aren’t reporting this data (Nebraska & North Dakota) and that needs to change. https://t.co/oe49yLPxKF
Organisations like the COVID Tracking Project, create feeds of key metrics to understand the trajectory of the pandemic and to add more context on charts of case data like test-positivity rates or the number of tests administered.
” Early on, this data was challenging if not impossible to source consistently at a state or county level,” Makulec reported.
According to Makulec, data visualisations of COVID-19 shift away from focusing only on visualising case and death data; for example, firms like Pentagram and Reuters Graphics have worked to create novel, different ways to tell the story of COVID that goes far beyond case curves.
“Visual simulations (like the coronavirus simulator illustrating the potential effects of social distancing in flattening the curve and the People of the Pandemic Game) have also played a key role in enabling understanding of complex epidemiology concepts in this challenging time,” Makulec explained. ” Games like People of the Pandemic also reinforce the impact of our individual choices on our communities through an interactive, cooperative experience.”
Ethics behind Visualising COVID-19 data
Not only do data researchers face the challenges of ensuring data quality, but they face significant ethical challenges in how people visualise data about COVID-19, how data is presented, and how to make sure the humanity and gravity of COVID-19 data remain in aggregate counts, bars, and lines.
Many data visualisations on social media spread without appropriate sourcing of the data or with misleading displays; for example, comparing COVID to the seasonal flu, which actually requires a more nuanced understanding of public health concepts and data.
Trump has persistently minimized the seriousness of COVID-19 by falsely comparing it to the seasonal flu. His denial and obfuscation squandered precious time, worsened the crisis, and cost lives. https://t.co/9HOqZcdZC4
” Social media allows people to speak but not necessarily to be heard because people can only consume limited messages per day. there are more special and influential messages delivered by politicians, and celebrities,” SMITH explained.
Visualising social networks would present comprehensive information and shorten the time people consume messages.
” Online conversations are networks, they have different shapes. The big circle is the governor’s account, and it generates lots of tweets, but each tweet in this circle is likely to directly connect to the governor’s account,” Smith said, explaining the graph.
“The idea here is to give people a tool to read hundreds-thousands of tweets per day, not by reading those tweets, but to abstract the essence of those messages.”
Move forward visualising COVID-19
The information itself is not visible or make sense sometimes, but good data visualisations make the invisible visible.
As consumers of data visualisations, we want to understand the data thoroughly, quickly, accurately, so we should support and encourage high-quality data visualisation and be responsible for what we share.
A multidisciplinary team of COVID-19 public dashboard project at the University of Sydney shows the cases by location in NSW, in collaboration with New South Wales Health and Australia Bureau of Statistics.
Yue Gao is a Master of Media Practice student at the University of Sydney. She is curious and has passion for life, justice, and equality in the world, UNSDG. Motto: "Live the life you love; love the life you live." TW: @kassiagao ; INS: @yue2993
Want fresh, thoughtful journalism in your inbox? Subscribe to Salience.
Salience showcases fresh, thoughtful journalism and creative works produced by students of the School of Art, Communication and English at the University of Sydney. Salience reflects the work of a globally oriented and engaged student body, with strong international student participation.
Subscribe to Salience
John Woolley Building A20 University of Sydney NSW 2006