How Coronavirus Could Have Spread On A College Campus

By Oliver Gladfelter

April 15, 2020

Last month, as coronavirus began to rapidly spread in the United States, many universities made the difficult decision to close their campuses and move classes online. Meanwhile, Liberty University’s president, Jerry Falwell, went a different route by actively playing down the threat of a pandemic and reopening campus. How much risk is a college president posing to their students by keeping them on campus?

Tap for next slide ➤

Say students were allowed to stay on campus, but classes were moved online, extracurriculars were cancelled, and parties were put on pause. Let’s also posit students mostly social distanced from others, but continued to see only their closest friends on campus. Could a virus still reach a significant portion of the student body? Let’s explore using actual friendship data I collected from my peers during my senior year.

Tap ➤

To be clear, I’m not an epidemiologist. But I do have some interesting data which might provide a fresh perspective about school closures during the coronavirus epidemic. When I was in college, I wrote for a data journalism platform called OneTwentySeven and regularly sent surveys to the student body to collect data for our articles. In February 2018, we asked students to list their best friends on campus (up to 5). So while there are hundreds of virus simulations out there, this one leverages actual social ties to visualize how a virus could hypothetically spread from patient zero to their friends, their friends’ friends, and so on.

Tap ➤

This network graph visualizes the survey data of Denison University students in 2018. The 1,400 dots represent students who either took our survey directly (600) or who were listed as best friends by survey respondents. These dots represent 64% of the student body. Each line represents a friendship between two students. You may notice the graph is remarkably connected - the “six degrees of separation” theory holds up pretty well here, with just a handful of exceptions. This undoubtedly would help foster the spread of a virus as we’ll see shortly.

➤

Meet The Class

Let’s get to know this student body a bit. Dots highlighted in green are all of the students who took my survey and were in a fraternity or sorority at the time. The gray dots are those who took the survey and weren’t involved in Greek life. The rest of the dots are students who were named as best friends but didn’t take the survey themselves, so we know less about them.

➤

How about class year? Here you can distinguish between seniors, juniors, sophomores, and freshmen.

➤

And here are the Democrats, Independents/Political-Nones, Republicans.

➤

Now let's meet some individual students. Here’s a freshman hailing from the West Coast. He identifies as an introvert.

➤

Here’s a junior from the East Coast. She’s in a sorority and identifies as a Democrat.

➤

And here’s me! Being from Ohio originally, I didn’t have to travel far to reach Denison. In February of 2018, I was a senior finishing up my political science major.

➤

Virus Simulation

Now that we’ve gotten familiar with the student body, let’s run a simulation testing how far a virus could reach once it arrives on campus. Let’s say I contracted a disease so contagious that it successfully infects everyone I interacted with most: my closest friends. 🔴︎

➤

After seeing and unknowingly infecting everyone I named as a best friend or who named me as one of theirs, there are now 7 additional cases on campus. The virus could stop spreading here if my friends and I formed a “closed network,” in which we each associated solely with one another. However, as the data show, each of my closest friends have their own friends whom they socialize with. So while I have just 7 friends in this simulation, I also have almost 100 friends of friends.

➤

Here’s how a super virus could theoretically spread through the student body, if there were no preventative measures or social distancing:

➤

The virus stops spreading after 13 days, once 1,266 individuals are infected - 90% of our graph. Remember, that’s assuming students weren’t going to class or extracurricular activities - just continuing to see their closest friends.

➤

That’s how it could have played out if I was patient zero. Would a virus have less impact if someone else contracted it first? See for yourself: click the button below to start a new simulation with a random patient zero or click on any dot to infect that student.

➤

Of course, no (known) virus is contagious enough to infect everyone it's exposed to. We know that the average coronavirus carrier spreads the virus to 2-3 other individuals, making it highly infectious, but not unavoidable. Use the Infectiousness slider below to determine how contagious you want to make the virus in this simulation. Now, red dots 🔴︎ still represent infected students, but cyan dots 🔴︎ are those who did not catch the virus, despite having an infected friend.

➤

Each cyan dot represents a student who not only evaded getting sick, but also avoided becoming carriers, effectively ending the domino effect of virus spread throughout many branches of the social network. This also helps visualize why social distancing is so powerful: imagine if these cyan dots didn’t get sick because they opted for video calls rather than going to their friends’ dorms. Removing just one person from a chain can protect dozens of others downstream. In other words...

➤

Social distancing works.

➤