The Illinois Institute of Technology now has a COVID-19 Confirmed Cases Dashboard, as first sent out in an email from Associate Vice President Hilary Hudson Hosek on October 30, 2020. That is a categorically good thing. It is better to have all of the available COVID-19 data in one place, as opposed to being spread out over weekly emails. Issue 4 of TechNews this semester, printed and distributed on Wednesday, September 30, led with a piece titled “No dashboard: Illinois Tech’s poor COVID-19 data communication,” and TechNews wishes to thank the university and its administration for fixing this error and implementing a dashboard. While snarky yet valid criticisms can be made about timing, this is not the article for that.
The dashboard is linked in weekly emails from the Associate Vice President Hilary Hudson Hosek, along with updated city and university guidelines, at the link https://www.iit.edu/COVID-19/confirmed-cases. As per the emails, TechNews strongly encourages readers to wear your mask covering both your nose and mouth, wash your hands, stay six feet away from others, and monitor your health.
The COVID-19 dashboard first presents a panel with three bar graphs for randomized testing data; since the week of September 21, 2020, Illinois Tech has been randomly testing between 150 and 199 members of the community for COVID-19 every two weeks and recording the data to keep an eye on the pandemic and its status on campus. The first graph represents the number of randomized tests per two week period, broken down by students, staff, and faculty. The second graph shows the number of positive cases from each two week period of randomized testing, also broken down by student, faculty, and staff. The third graph from this section shows the total positivity rate from each two week period of randomized testing. This most recent period recorded, starting November 2, was the worst so far, reporting one faculty and two staff cases out of 155 tests for a 1.94 percent positivity rate.
The second panel of the COVID-19 dashboard presents bar graphs for daily totals. The first bar graph in this panel presents daily positive tests, as reported to the university, broken down by student, faculty, staff, and contractor. The second bar graph in this panel presents the total number of daily tests reported to the university, as broken down by student, faculty, staff, and contractor. Large numbers of tests were conducted by the school or reported to the university starting August 17, around the start of the semester, peaking with over 300 tests in one day on August 20. This number waned, until the university began conducting its randomized testing every two weeks, with testing values of over 150 now appearing on September 21, October 5, October 19, and a smaller value of just over 100 on November 2. The second panel has no mention of a positivity rate.
Again, it is absolutely a good thing that this dashboard has been made by this university. However, statistics are notoriously hard to get to tell the “truth” about any situation, and this dashboard is no exception. While important and good by the university, it deserves analysis.
It makes sense that the university would lead with its first panel of randomized testing, which the university itself conducts. This is the only section which includes a positivity rate, which only includes random testing data. There are problems when constructing a positivity rate from non-random data, though, such as students potentially getting tested at third-party facilities and not reporting negative tests to the university, or the outsized effects that positive tests from contact tracing through a small outbreak can have on an overall positivity rate. With not including non-random positive tests, though, the verdict from the random sample positivity rate can be misleading as it relates to the rest of campus.
The second panel — with graphs for positive tests and total tests per day — is also open to criticism, as showing data per week instead of per day could be easier to digest, at the tradeoff of showing larger numbers of positive cases at a time, which the university could be hesitant to do. One week of one case per day, spread over a graph which reports by day, may potentially look better than showing a greater value of cases (seven) per week. There is also no mention of
There is a fine balance between including too little information, which provides an incomplete picture, and including too much information, which crowds a graphic, however comprehensive. Again, we appreciate the hard work the university has put into this graphic, even if not all aspects are perfect; not all aspects need to be perfect to appreciate improvement.
As of the time of writing, the COVID-19 pandemic is worsening in the United States, with upwards of 175,000 new cases per day and total deaths approaching 250,000. Illinois is now reporting 15,000 new cases a day with a death total over 11,000. The United States has an estimated population of 328 million, with Illinois having an estimated population of 12.6 million. The Illinois Tech COVID-19 dashboard is now reporting a total of 100 cases, between students, faculty, staff, and contractors.