That @Nytimes op-ed in which @fmanjoo counts 100+ contacts outside his bubble, then decided he loves his family so much that he'll expose them to risk w/ a Thanksgiving gathering anyway? It gets worse: for many families, 100 is a gross underestimate.
1/n https://twitter.com/nytopinion/status/1329782395369902081

College students around the country are returning home for Thanksgiving. Many are coming from unis that tried to keep the community safe(r) with a bunch of adaptations, including shifting to some mix of in-person and online classes. (78% of unis, per @chronicle.)

2/n

The shift to mix of on-line and F2F instruction changed the structure of co-enrollment networks -- the network of students connected to each other thru in-person classes.

Ben Cornwell & I wrote about the pre-pandemic network @Cornell here: https://sociologicalscience.com/articles-v7-9-222/
3/n

With @barum_park, Ben Cornwell and I updated these analyses using enrollment data from Fall 2020, when about 70% of Cornell's class meetings (inc. disc, labs, etc) are on-line and about 30% in person. Paper isn't published yet, but here's a sneak peek of results. 5/n

In Fall 2020, the F2F enrollment network encompassed fewer students: ≈50% of Cornell's 23K enrolled students. Connectivity in network declined from F2019: the average path length connecting pairs of students grew & a smaller % of students are connected by multiple paths.
4/n

In Fall 2020, a larger share of student ties in F2F network are between students in the same field, likely because large (50+ student) gateway courses that serve multiple majors or general ed requirements went online. Some depts. may have reserved classroom spots for majors. 5/n

The "small world" network of enrollment face-to-face classes became a bit larger. Great news for Thanksgiving, right?

Here's the thing: a college student's homecoming is still indirectly connecting family and friends outside the household bubble to 100s -plural - of others.
6/n

The average Cornell student taking at least one F2F class is directly connected to 65 other students through classes. This is much less than 2019 average of 529 students, but it's still a lot. And, this count doesn't include connections thru dorms, sports, parties, etc.
7/n

What about indirect contacts? In Fall 2020 F2F enrollment network, 10% of student pairs are connected through two steps or fewer and 55% through three steps or fewer. This is way down from F2019 (59% & 92%), but it STILL means G-ma is 3 degrees of separation from 100s. Plural.

As any professor can tell you, co-enrollment is not the same thing as co-attendance. Co-enrollment in a class also doesn't mean close physical contact, although it does entail sitting in same room for 50-145 minutes at a time (even if at 6'+ distance).
9/n

We shouldn't treat college students as lepers. At @Cornell, the vast majority have been absolutely amazing about following public health recommendations to protect the campus, the community, and their semi-residential term. Statistically, few will be infectious. 10/n

Still, the return of college students for T-day break is rapidly, even if temporarily and unintentionally, expanding the number of indirect ties a focal household has to potentially infectious others. 11/n

Ideally, college students would test negative before traveling home, quarantine for 2 weeks, and then -- and only then -- rejoin the family bubble, just in time to sit down for an early December Thanksgiving dinner.
12/n

That's unrealistic for many reasons, not the least of which is that many college students desperately need hugs from parents (and vice versa) after a bizarre, stressful, and isolating fall term. But families in same household can still show love by NOT exposing others.
/fin