Corona virus presentation by the Dutch CDC, also some thoughts on the audience for these sorts of presentations

Anne Pier Salverda writes:

I’ve attached a corona virus presentation by the RIVM, the Dutch equivalent of the CDC. This briefing is happening as I send this email.

The presentation is comprehensive and comprehensible, and it hits all the marks in the data visualization and communication department.

Are you aware of any comparable presentations by the CDC?

It’s a 20-page document with lots of pretty pictures. My Dutch isn’t so great anymore but I can still follow most of what’s going on. I guess maybe the CDC is preparing something similar.

It’s funny—we think of this sort of thing as education for the general public, but it’s really just as important to be educating the subset of the public that has control of major resources: I’m thinking here of government officials, media figures, corporate officers, maybe labor leaders too (although I guess they’re less powerful than they used to be). The job of the CDC is in part to inform the resource allocators and decision makers who can then get their act together and do something. From that point of view, perhaps the biggest communication contribution from the CDC is not to have someone stand next to public officials in a press conference, but rather to prepare high-quality and convincing documents that can persuade the powerful.

5 thoughts on “Corona virus presentation by the Dutch CDC, also some thoughts on the audience for these sorts of presentations

  1. Some context if useful:
    This presentation was given this afternoon to the houses of parliament (at least selected members from the different parties). With follow-on extensive Q&A.

    The modelled estimates of the effective current reproductive number give some hope that the measures are working. However the uncertainties are still very big (e.g. those modeled in the plots with all the overlaid lines). The disease may still exceed ICU capacity.

    In the presentation there was also a lot said about the need for more reliable data about the actual spread of the disease in the population. Trough randomised samples testing for the presence antibodies. Also by studying children – who luckily seem to have the disease less with fewer symptoms – but for which it is unclear whether they could still play a (big/small) role in transmission.

    • This is indeed a very detailed presentation. However, I wonder if it places too much confidence on a particular model and uses counterfactual estimates from that model to inform policy. In particular, it recommends against total lockdown because this may lead the number of cases to spike at a later date. A spike in the number of cases after a total lockdown has never been observed before. Combine this with the fact that there appear to be a lot of measurement error, some of it systematic. For example, one can only get tested after calling in sick and after that you need to wait before they come for you. This can lead to a lot of biases in the data: (i) some people may be reluctant to call, or may not be able to express the seriousness of their condition, (ii) human factors may lead to some calls not getting registered, and (iii) most importantly, supply side issues restrict the number of tests you can do in a day.

      So I am a bit skeptical about their model counterfacuals, and I would be happy to hear what Andrews thoughts on this, especially given that he has emphasized measurement error in earlier posts.

  2. They should hire Francis Nightengale in that case. As far as I know, how to best communicate estimates with uncertainty to policy makers still remains understudied relative to lay audiences. Probably in part due to what samples are convenient. Baruch Fischoff and Anne Bostrom come to mind as researchers in uncertainty communication who have looked more specifically at policy makers and other decision makers. I suspect (and have at times observed) that in government reporting especially, its commonly the case that the researchers prepare reports for the policy makers, then the policy makers communicate with the public, but I haven’t come across much research that discusses this pipeline.

    Of covid-19 articles making the rounds, this one by Tomas Pueyo got me thinking a bit about whether the right rhetorical strategy for this crowd might be to remind them of their power and convince them of the opportunity to distinguish themselves as great leaders: https://medium.com/@tomaspueyo/coronavirus-act-today-or-people-will-die-f4d3d9cd99ca. Though I would certainly hope that regardless of the appeals to ego, the more power the decision maker has, the greater the need for comprehensive reporting, discussion of assumptions, and other forthright communication of uncertainty

Leave a Reply to Jessica R. Hullman Cancel reply

Your email address will not be published. Required fields are marked *