key: cord-1036645-pe6br42a
authors: Dean, Natalie
title: Statistical successes and failures during the COVID‐19 pandemic: Comments on Ellenberg and Morris
date: 2021-05-08
journal: Stat Med
DOI: 10.1002/sim.8934
sha: 6eb079da64c4ea33a97d7b024354d64ca745d0f0
doc_id: 1036645
cord_uid: pe6br42a

nan

As statisticians, we must be advocates for data quality and accessibility. The best treatment for these data woes is prevention, and statisticians have a key role in highlighting why we need high-quality data to guide the response. This includes demanding more transparency in the methods used to report and calculate statistics, which is crucial for making comparisons across locations. Even something as simple as the total number of SARS-CoV-2 tests conducted can be very complicated, as the COVID Tracking Project has described in their blog. 7 Given their deep understanding of COVID data, the COVID Tracking Project group has also offered public recommendations for a national dashboard and tracking vaccinations. 8 We can also be advocates for making data easy to access. For many states, data have not been consistently available in machine-readable format to enable automated retrieval. Statisticians can highlight the importance of open data to encourage the many independent data analysis and modeling efforts that have yielded useful insights.

A Work in Progress-Statisticians and the Media. A reporter recently commented to me, "SCIENCE IS THE NEWS NOW." Given the dramatic and persistent impacts of this pandemic on daily life, new scientific insights are constantly in demand. Of course statisticians can contribute to the cause through their scholarly work-be it designing a clinical trial, analyzing surveillance data, or modeling disease dynamics-but there are other opportunities to have a wider impact. The landscape is moving so quickly that peer-reviewed journals are straining to keep up, and so science is happening in pre-prints, on dashboards, in opinion pieces, and even on social media. To maintain quality, more eyes are needed for informal peer review. As statisticians, we can offer a range of skills to improve public understanding and correct misconceptions. As a small example, I have now taught the concepts of test sensitivity, specificity, and positive predictive value to my students and at least a dozen reporters with questions about antibody testing.

Unlike the early days of the HIV pandemic, there are many more avenues to engage with the public. These efforts fall under the heading of service, which does not always get its due in academic circles, but as a community we should acknowledge the immense value this work adds to the public discussion. The time required can range from modest, for example, a 15-minute chat with a journalist who will then translate your insights into a larger story, to more intense, for example, maintaining a weekly blog. Social media is a particularly accessible way to join the conversation. Scientists can stay up to date on the latest research and layer on top of these their own interpretations. Journalists are very active on social media, and they may reach out for comment if they find your insights to be uniquely informative. Of course media training is not a part of our curriculum, and some statisticians may be uncomfortable talking to the press. A blog post entitled "Top tips to statisticians communicating through the media, especially in the time of COVID-19" offers advice, an act of service in itself. 9 The helpful tips include "don't be lured out of your comfort zone" and "keep off the (statistical) jargon." Adding my own, as statisticians we may be most intrigued by the details, but reporters are most interested in the big picture. Learn how to strike a balance.

As noted by Ellenberg and Morris, during the COVID-19 and HIV pandemics, "[s]tatisticians and other data scientists with expertise in integrating and synthesizing information across multiple data types … play a critical role in proper interpretation of these data." We are able to bring nuance and insight to an often over-simplified public discussion, where an out of context point estimate can turn into headline and then into a media frenzy. We can enter into conversations early to encourage high quality study designs, and support sound statistical inference along the way. Yet our role can also extend far beyond data analysts. Whether your research focuses on infectious diseases, genetics, air pollution, neuroimaging or another area, statisticians can also be scientific leaders. Statisticians can testify in front of governments. By viewing ourselves as well-rounded scientists, we are able to achieve an even greater level of impact. DATA AVAILABILITY STATEMENT Data sharing is not applicable to this article as no new data were created or analyzed in this study.

Natalie Dean https://orcid.org/0000-0003-3884-0921

COVID-19 Prevention Network. COVID-19 Prevention Network: about us

HIV Vaccine Trials Network. Scientific achievements

Early detection of Covid-19 through a citywide pandemic surveillance platform

REMAP-CAP response to the COVID-19 pandemic

Covid-19 in the US: we're not getting full value from our data

Counting COVID-19 tests: how states do it, how we do it, and what's changing

How we hope vaccines will be tracked

How to cite this article: Dean N. Statistical successes and failures during the COVID-19 pandemic: Comments on Ellenberg and Morris