Ranking Virginia Colleges for COVID Performance, Round II

by James A. Bacon

It is a legitimate question to ask: Which of Virginia’s colleges and universities are doing the best and worst job of managing the COVID-19 epidemic? Over the weekend, I posted some numbers showing that Virginia Tech, James Madison University, and the University of Virginia had posted the largest numbers of confirmed COVID cases in the state. But the raw numbers don’t tell us much by themselves. As large public universities, those three institutions have among the biggest student bodies. Perhaps it would be more appropriate to compare the colleges after adjusting for the number of students enrolled. That’s what I’ve done here.

In the table I have expressed the number of confirmed COVID-19 cases, as reported by the New York Times last week, as a percentage of average full-time-equivalent enrollment, as reported by the State Council of Higher Education for Virginia (SCHEV). Let me be clear: The resulting number is not a percentage of the study body that has been confirmed to have COVID. The confirmed-case numbers also include faculty and staff. At UVa students comprised a large majority of COVID cases but not all. The same pattern likely holds true at other institutions. Consider the percentage as a rough relative measure of the prevalence of the virus at each campus.

This table reveals that the incidence of COVID-19 was the worst at Hampden-Sydney College in Farmville. (By way of comparison Longwood University, also located in Farmville, experienced less than one third the number of confirmed cases.)

The table also shows that JMU had the second-highest incidence of confirmed cases. Because JMU is more than 20 times larger than Hampden-Sydney, the impact was far more profound.

Other observations worth noting:

  • UVa appears 9th highest on the list of 35 institutions, seemingly but not necessarily at odds with President Ryan’s portrayal of the university’s response as heroic.
  • Virginia’s historically black colleges and universities — Norfolk State University, Virginia State University, and Hampton University — stand out for the extremely low incidence of confirmed cases.
  • Eastern Virginia Medical School — a medical school! — was in the middle of the pack for COVID cases. Wouldn’t one expect doctors and aspiring doctors to be more conscientious about preventing the spread? Or do we cut them slack because they were more likely to be exposed?

Though more refined than the previous table I published, one should be careful about using this data to grade how effectively college administrations responded to the COVID challenge. One factor to consider: How aggressively did the various institutions test for the virus? The more tests they administered, the more confirmed cases they identified but the more effectively they could combat the spread.

In the case of UVa, for instance, the university ramped up its testing capabilities in concert with a contact tracing program and the setting aside of residential buildings for isolation and quarantine. One cannot rule out the possibility that the high number of confirmed cases there reflects a more vigorous virus-fighting strategy. Conversely, perhaps institutions with low numbers just weren’t testing. This information is crucial for reaching a judgment.

Another factor is how committed the institutions were to maintaining in-person instruction. Each institution rolled out a unique blend of online, hybrid and in-person classes. If one university conducted most classes online, one would expect it to experience fewer infections than a university that taught more in-person classes. One must consider the tradeoffs between fighting the virus and the quality of instruction.

The purpose of publishing this data is not to condemn institutions that have a higher incidence of infection or praise those with a low rate. The purpose is to provide students, parents, and boards of visitors a starting point for asking tough questions and holding administrators accountable for performance. In UVa’s case, Ryan bragged about the great job his team had done without providing any context, and the Board of Visitors complimented him for a job well done without asking for any. University stakeholders should demand more.

There are currently no comments highlighted.

25 responses to “Ranking Virginia Colleges for COVID Performance, Round II

  1. This is good but it has strong potential to devolve into a naval-gazing info swamp.

  2. Now the question is, which were all on-line, which were holding mostly in-person classes, and were any a mix?

    Just saw some data reported by one of Richmond’s major private schools for the fall, one which is mostly in-person. Zero 0n-campus cases for students and staff in the lower and middle school, a bad start for September for the upper school (you may remember the coverage) but then even it settled down to a handful of cases per month. What transmission they know of was elsewhere, not on campus. But college kids are clearly at an age where the disease is more prevalent.

    • Online education works against this disease and is an effective means of education. I’ve not been exposed to covid and have been taking online gynecology courses for 30-some years.

  3. re: college kids and college towns and off-campus congregating.

    that’s why this exercise has the potential for a data swamp.

    Do all this slicing and dicing and at the end – you’re gonna get an “uh huh”.

  4. Liberty University is largely on-line. On-campus enrollment is about 14,500. This would put it around 7.2%. A little higher than UVa & Tech, but not quite as bad as the outcry predicted when they reopened the dorms last spring.

    • On campus enrollment is a reasonable metric, more reasonable than looking at total enrollment is some/many are not actually on campus – in class or dorms.

  5. From today’s VP/DP…
    https://www.pilotonline.com/opinion/vp-ed-cartoon-a-1214-20201213-iewtp37st5crpejhb6ypq4hf6e-story.html

    Now, I ain’t an anti-vaxer, but if the stuff glows in the dark, I ain’t takin’ it!

  6. As I recall from my doctor friends, some med students spend time in hospitals as part of their course of study. That would make them part-time health care workers and would explain a higher than expected COVID rate.

  7. Just a note… community spread don’t stop at the wavy walls.

    HU, all the way down at the bottom is largely separated from Hampton by a river. Sorry, rivah.

  8. Interesting and telling table, thank you for sharing. If you continue exploring, I have two requests:
    Report student cases rather than student + staff (or include staff in population figures)
    Report percentage of students tested (more tests will presumably reveal more cases; one could

    Mason claims 229 student cases https://www2.gmu.edu/campus-covid-19-data

    • FWIW: Fine print on UVa dashboard: UVA testing prioritizes individuals who are symptomatic or close contacts of infected persons. Tests taken outside the university are not reflected in this tracker.
      Students only 1124
      Faculty, staff and contract employees: 230

  9. Regarding James Madison University…. a correspondent reminds me that the numbers of confirmed cases — 1,697 understates how badly the COVID-19 response was bungled. President Alger sent thousands of students home for six weeks or longer, after which hundreds chose not to return to campus.

  10. Not the only bungling. VDH has 51 outbreaks in colleges or universities with a total of 2,824 cases. I had a total of 3,329 for 4 universities with COVID dashboards on Sept. 25. Jim’s list is over 9,400.

    VDH said whether the university cases get recorded depends on where the testing was done and whether the reporting lab used the student home addresses or university address for positive results. If an out of state address was given, it wouldn’t be counted. If tested at the school, then it was up to them to report to VDH. VDH was not checking with schools.

    • You all have no idea what you’re talking about. And this story is completely inaccurate.

      If you go to the university COVID dashboards and actually carefully read them, you will see that JMU is the ONLY university in the commonwealth reporting both positive cases tested in its health center AND student self reports from tests off-campus. The other universities are only reporting positives cases tested in their health centers.

      So if you are actually comparing apples to apples, here’s the current reality:
      Virginia Tech – 1,721
      UVA – 1,410
      JMU – 763

      JMU made the decision to report student self-reports out of an abundance of care for our local community.

      So the real bungling is this lousy reporting.

      • Excellent point. Before we can make legitimate comparisons, we need to know how the universities are collecting and reporting their data.

        • Now that you know the data used in the chart are not comparable, are you going to change it? If you use only students tested positive at the JMU health center – as the other universities are reporting – that would put JMU at 3.60.

      • Conclusion:

        1/ Only JMU has been honestly reporting Covid-19 unfavorable facts to its students, and the public.

        Meanwhile,

        2/ UVA and Virginia Tech were intentionally hiding unfavorable Covid – 19.

        In plain talk, UVA and Virginia Tech have been lying all along by intentionally hiding key Covid 19 facts that would show the real picture of what was going on among their students. Frankly, I am not surprised. This conduct increasingly seems to be standard operating procedure for these two institutions. Simply put, one cannot believe a word these institutions say based on trust. Particularly so because, at least in the case of UVA, its leaders not only hide key facts, those leaders hold public press briefings, making claims for themselves, that those leaders know to be demonstratively false and misleading. This is malicious conduct engaged in for private gain and public expense, a game UVA has been playing since 2012 at least.

        • That’s one way to look at it. Another way to look at it is that student self-reported positive tests off-campus are not verifiable. So there is a data integrity question.

  11. sorry folks…I’m a day late and a dollar short on this discussion…I’m an HSC grad…long time ago…I spoke to someone at HSC regarding their ‘numbers’…
    BTW, they have a website ‘dashboard’ regarding their numbers. Bottom line,
    of the 128 ‘cases’, 104 were freshman that the college tested BEFORE they were allowed to move into their dorm rooms. The students were told to return home. Sometimes the numbers reported do not reflect the context of the reports. Not only in this case but in many others…Be careful what you wish for when you look at ‘data’…

    • Not to doubt you, but how reliable is your source? Was this person in a position to know, or just spread a rumor? It’s important, you see, for this is exactly how rumor becomes fake news.

    • Another good point — to make a fair comparison, it is helpful to know if students were infected by the virus on campus or if they brought it from home.

      • This. In spades. Some schools (and JMU is very notable in this regard) did NOT, for some mind-boggling reason, employ “intake” testing prior to their students returning to campus this past Fall. Yet another reason why using the NYT data as a cross-institution comparison tool is problematic… as the NYT itself explicitly notes at several points in the accompanying article.

        • It’s not a stable environment for data. It’s coming from a lot of different places, at different times, and may not represent the same status, etc.

          “testing” has not been consistent either. Some test. some don’t. Some do it frequently. Others less so.

          testing only tells us at one point in time whether someone if infected or not. A week later, they may test the opposite of the first time.

          If Higher Ed tests on “intake”, do they also rest when students come back from break?

          On top of all that – the tests are not cheap and they are not perfect either. Different tests have different accuracy in terms of false-positive or false-negative.

Leave a Reply