As the new coronavirus has spread around the world, many people have begun to rely on online maps to understand it. Researchers at Johns Hopkins University used a software package called GIS to create an interactive dashboard with a map, numerical data and charts. The New York Times has presented updated information about where cases appear in two types of non-interactive maps: a world map with the number of cases written over each country, and maps of the United States, East Asia and Europe with proportional circles over provinces and cities.

It’s not only ordinary citizens who rely on maps like this. Acting deputy homeland security secretary Ken Cuccinelli asked Twitter for help when he briefly wasn’t able to view the Johns Hopkins map, and complained about the institution’s bad timing in limiting access.

This crisis isn’t the first time that mapping has played an important role in tracking the spread of a disease. A celebrated map of a cholera outbreak in 1850s London played an important (though probably exaggerated) role in changing how Britain dealt with epidemics.

But maps — like every other way of presenting complex information — have drawbacks as well as benefits. Inevitably, they highlight some aspects of the situation and underplay others. Every map embodies an argument. Here is what you need to know about mapping coronavirus.

Maps do more than present information

Mapmaking has been transformed by Geographic Information Systems (GIS) software, which makes it very easy to create your own map of a phenomenon, with data and a little bit of technical ingenuity. However, mapmakers’ assumptions, biases or omissions may unintentionally become embedded in GIS-produced maps. This isn’t to say that all maps “lie,” but all maps emphasize certain things and leave others out — intentionally and unintentionally. Mapmaking choices have shaped the politics of issues such as the nation-state, colonialism and boundary negotiations.

The complications involved in mapping coronavirus

The Johns Hopkins and New York Times maps, and others like them, show how mapmakers’ choices can have unexpected consequences. The Johns Hopkins map illustrates the capabilities of GIS-based online mapping. The map shows cases by location; users can pan, zoom and bring up detailed data; and it contains extensive information about sources. As a proportional symbol map, it uses circles to represent the number of cases for each place: an enormous circle over Hubei province, large circles on South Korea, Iran and Italy, and so on. The New York Times maps are more static, but also encapsulate a lot of valuable information.

However, the ways in which they present information has consequences. Obviously, the fact that they are maps — rather than lists, charts or graphs — emphasizes the importance of physical space, suggesting that the outbreak is spreading more to nearby countries than to distant places. Since the virus has moved via air travel, the implicit emphasis on proximity can be misleading (though other maps have added air travel information). By using a default basemap that just shows state boundaries and few geographic features, they do not capture important information about such things as whether the borders are open or closed.

Their choices of color and centering also tell political stories. Many (but not all) coronavirus maps use shades of red and orange, colors associated with danger in many cultures. This encourages fear even as most official statements are urging calm. Mapmakers often wish to place their own location near the center and at the top, but in the case of coronavirus, the Chinese government is unlikely to be happy with how many maps default to centering on China.

The Johns Hopkins and New York Times maps choose to use proportional circles on the map to show the size of contagion. This is a reasonable choice, and other ways of getting this information across have their own problems. But it can easily be misread. People might think that the circles show the areas of effect rather than the number of people affected. The circle for Hubei province, for example, is so large on the New York Times map that it extends over most neighboring provinces as well. If the user zooms out on the Johns Hopkins map, it covers much of East Asia. Experts will know that the circle indicates a number pinned to a location, but many viewers might initially think it is a sign of geographic spread.

Finally, the data comes largely from governments and international organizations and may unintentionally reflect their political interests. For example, the Johns Hopkins map locates the repatriated U.S. cases from the Diamond Princess cruise ship where the ship is docked in Japan, not where those individuals are currently being quarantined or treated. They describe this depiction as “consistent with the” Centers for Disease Control and Prevention. Certainly it suggests that the situation is being well enough managed by the CDC that the public need not worry about those cases’ current locations.

Read maps carefully

Mapmakers are discussing how best to map the current outbreak. It’s hard to present information on something so fast-moving and complex to a wide audience. One suggestion is that it might actually be better to “simply use a table or a graph instead of a map” to get the relevant information across. When you are reading a map of coronavirus, you need to be cautious about what it is saying, explicitly or implicitly. It is unlikely that the mapmaker is trying to trick you, but maps have hidden biases and limitations. Because of their “God’s eye view,” maps more than charts or texts may make readers think that they have complete knowledge, but this is an illusion.

Jordan Branch, currently assistant professor of political science at Brown University, will be assistant professor of government at Claremont McKenna College starting in July.