Post #1027: A correction on my William and Mary calculation

Posted on February 24, 2021

Same story as before, just a different baseline.  The upshot is that the red line — the expected count of new COVID-19 cases — should be based on 70% of total student enrollment.

And as a result of that correction, W&M’s COVID-19 incidence is not much different from what you would expect, based on the community rate.  But it’s not (yet) better than you would expect.  At the current rate, it’s going to take another week or two to get there.

Details follow.

When I constructed the graph above, I initially based it on the entire William and Mary enrollment.  Figuring that almost all students would return to campus, and that only a small portion of students would opt for remote learning.

I check the William and Mary COVID-19 dashboard periodically, and I’ve noticed that the number of tested students figure has stopped rising.  It has settled at around 6500, or about 70% of the total W&M enrollment.

And so, my comparison number should have been based on 70% of total enrollment.   That’s the red line above, which I have now recalculated to be 70% of what I was showing before.

The fact that the W&M COVID-19 dashboard only reflects 70% of total W&M enrollment does not appear to be due to students opting for remote learning.  Instead, I think this is a reflection of the proportion of students that live on the Williamsburg campus.  The W&M COVID-19 numbers are for students at or near the Willamsburg campus, and only about 70% of total W&M enrollment lives on the Williamsburg campus.

The upshot is that the story remains the same — nothing is out-of-control.  But W&M has not yet gotten cumulative total cases below what you would expect, based on the rate of diagnosed new cases for 21-to-30-year-olds in Virginia.  Give it another couple of weeks, at the current rate.

I guess that’s not surprising, given that they test 100% of the population, and re-test frequently.  I’m just embarrassed to have made that mistake.  I assumed that the W&M number was for all of W&M, and it is not.