Effective Data Visualization In the Era of COVID-19
Stanford University Associate Professor Kristin Sainani recently conducted a webinar to discuss the role that data visualization plays in educating the population and sharing information during the coronavirus crisis. In particular, she highlighted some of the most common visualization issues that can send the wrong message.
Choose the right type of graph for your presentation
The average person may not know the difference between a bar graph and a line graph, but the distinction between different types of charts can have a huge impact on the clarity of your data.
Professor Sainani called out a few examples where the wrong type of graph was used to display information. In one instance, the designer used a bar graph to show how world record times for the 100 meter dash has changed over the years. The intervals between these different times were so minute as to be virtually indistinguishable. Plotting the same data out on a line graph, Professor Sainani was able to visualize the information in a way that drove home the fact that world record times have dropped in recent years.
Keep your charts proportional
In a perfect world, there would be no such thing as a misleading data chart. But, here in the real world, deceptive or confusing graphs are extremely common. The main culprit? Not keeping the graphical presentation proportional to the numerical value of the data.
Professor Sainani showcased a recent example of this phenomenon, courtesy of CNN. The media outlet presented a bar graph visualizing COVID-19 cases in the U.S., Italy and China. A quick glance at the chart would give the impression that China had roughly twice as many cases as the U.S. or Italy. In fact, according to the figures listed on the chart, China had only about 1,000 more confirmed cases than the U.S.
Keep your charts and your message clear
Chart and graph designers can make a number of mistakes that detract from the message they’re trying to convey. For instance, using the wrong color scheme could make it difficult for the audience to easily discern the scale of difference between various subjects.
Cluttering the screen with too many data points and visualizations is another common misstep. If you overload your chart with too much information or too many graphical elements, there isn’t a single component to draw in the viewer’s attention.
In some cases, graph creators don’t appear to have a take-home message to communicate in the first place. Every design decision should be driven by an underlying message. Without it, the resulting chart will feature a smattering of data points that are relatively meaningless. The most effective visualizations tell a story, and that starts with a core idea.
Of course, every data visualization should be easy on the eyes to draw in viewers and appeal to their senses. Especially now during the coronavirus crisis, people rely on charts to get a better understanding of the pandemic and how the situation is developing. Creating effective data visualizations will help disseminate accurate information with the right message.
Watch the full webinar here to see more examples of data visualization and hear Professor Sainani’s thoughts on best practices to follow. If you want to take a deeper dive into statistics, Stanford has several statistical analysis courses focused on different fields. For instance, the Medical Statistics Certificate program can teach you everything you need to know about the techniques commonly used in medical research. Enroll today to start honing your analytical skillset with the help of industry experts and real-world examples.