‘A Very Small World’: How Data on Student Enrollment Could Help Colleges Stop Coronavirus’s Spread

If everything returns to normal this fall, the average Cornell University undergraduate will come into contact with more than 500 students in a typical week, according to a working paper by the sociologists Kim Weeden and Benjamin Cornwell. Freshmen and sophomores, who tend to take larger classes, will cross paths with more students; for juniors and seniors, a little less. And that’s just in class. It doesn’t include dorms, gyms, or cafeterias.

If a student were to get sick with Covid-19, hundreds of people could be exposed to it before that student showed any symptoms. Cornell would very likely have to close its campus immediately.

“What did surprise me,” Weeden, chair of the sociology department at Cornell, said, “is the extent to which it really is a very small world.”

Last month the spread of the coronavirus forced classes online and sent students and faculty members home. College campuses, places that are designed to bring people together to learn and develop ideas, had become a potential hotbed for a dangerous disease. Now that governors are talking about a leveling of new coronavirus cases, colleges are laying off staff members, and record numbers of unemployment claims are being filed, campus leaders will need to devise plans to reopen their classrooms.

A small number of researchers, like Weeden, who study and analyze social networks, think they can help. They’re turning to data about students and what courses they take — data collected in some cases to encourage more and deeper connections — and using it to map the risks of bringing students back together.

Colleges and universities were “built to tightly connect a large community of individuals so each can learn from every other with as little effort as possible,” said Timothy A. McKay, associate dean for undergraduate education at the University of Michigan at Ann Arbor. McKay has done similar research and is in contact with Weeden and Cornwell.

“Such a small-world community is great for spreading ideas,” he said. “It’s also, in some ways, great for spreading a contagious disease.”

Weeden and Cornwell’s recent research is the product of some quirks of fate. The data they used was collected three years ago as part of a project to analyze distribution requirements at Cornell. They looked at anonymized information that showed what courses students were enrolled in by their major. Weeden usually researches inequities in higher education. Before the pandemic, she was studying Department of Education data that was restricted to a particular computer in a particular room on the campus. She can’t go to that room now — she’s sequestered in a home office with two cats and an exercise bike — so she turned her attention to this project.

The work is in a very early stage, and the paper Weeden and Cornwell have written is not yet peer-reviewed. As a sociologist rather than an epidemiologist, Weeden refused to predict exactly how the virus would move through the campus. But this is a start, she said, and she hopes it will spur other research. It’s already gotten a lot of attention.

Weeden posted a graph showing her findings five days ago on Twitter. It was retweeted more than a thousand times. University administrators, parents of college students, and other researchers who are looking at similar data on their own campuses responded to it. They all wanted to know what it meant college leaders should do.

Should universities resume face-to-face instruction in fall? Ben Cornwell and I posted a working paper with relevant evidence from @Cornell on the structure of enrollment networks that connect students and classes.

Summary in thread, preprint here: https://t.co/JQaudyAAoK
1/11 pic.twitter.com/3bwF3lJU65

— Kim Weeden (@WeedenKim) April 12, 2020

Answers in the Data

Weeden says she doesn’t know. But the data does have some answers. She and Cornwell wanted to see what would happen if their university canceled classes, or put classes online, if they had more than 100 students. Cornell has a very popular course on wine, for example. It brings together students in different majors. When those students leave the wine class, they spread out across campus to economics, engineering, and English classes. If the university just didn’t offer that class and others like it for a semester, the researchers wondered, would that break up the network and potentially slow the spread of the disease?

Not really, was the disappointing answer. But, Weeden said, if some large courses moved online, that might free up space that would allow smaller classes to meet in large rooms. Students could then sit more than six feet apart, reducing the potential for the spread of the disease. That could have an effect. As she said, more research is needed.

U. of Michigan

That’s where McKay comes in. The Michigan administrator is also a physicist, and when he saw Weeden’s tweet, he suggested a collaboration. One of his Ph.D. students, Uriah Israel, recently published his thesis on connectedness on campus. On Thursday morning, Israel, McKay, Cornwell, and several other researchers at both universities gathered on a Zoom call to share notes.

Israel had spent the last few weeks at home, preparing to start a postdoc on machine learning and single-cell biology at the California Institute of Technology. He had no idea that his research might be used to plan how to reopen universities.

“This has been pretty weird,” he said.

His work was meant to map how students interact with one another through classes in order to foster the spread of ideas on campus by making those connections stronger and more frequent.

Coronavirus Hits Campus

As colleges and universities have struggled to devise policies to respond to the quickly evolving situation, here are links to The Chronicle’s key coverage of how this worldwide health crisis is affecting campuses.

• Here’s Our List of Colleges’ Reopening Models
• Will ChatGPT Change How Professors Assess Learning?
• Students’ Trust in Their Colleges Held Steady During Covid’s Early Days, Study Finds

“Students could use this information to seek out courses which will connect them to a more diverse array of other students,” he wrote. “Faculty members might use these networks of connection to better understand where their students are coming from and where they might go.”

Now his work might be used to limit where students are coming from and where they might go.

“The trade-off that people are considering these days is the possibility of not coming to campus,” McKay said. “Or perhaps not even opening your campus.”

Despite Weeden’s finding that canceling large classes would have a limited effect, McKay could imagine keeping some classes online — such as large lectures and courses that bring together people in different majors — as a way to shrink the network of people a typical student comes into contact with. That would make it easier to isolate groups of people if there is an outbreak.

“We don’t know all the answers yet,” he said.

But time isn’t slowing down. Answers or no answers, college and university leaders have about four to six weeks to figure out what the fall semester will look like.