david canning

Wave 2 of our COVID-19 online survey now available! Even if you took it before today you can now take it again so we can update! hsph.harvard.edu/pgda/

People have been asking for the power calculations for a random sample - how the sample size and affects the confidence intervals - some preliminary work on my website hsph.harvard.edu/david-canning/

@mlipsitch Thanks Marc great to have your support on this - people keep asking why it is not happening already

Like just about everything in COVID, my reading is about a week behind (well, that's better than many things in this pandemic). This is a really good idea bostonglobe.com/2020/04/15/opi… though getting consent from those people to be tested might be a challenge. But best idea I've seen.

Great study on false positive/false negative rates on COVID antibody tests -really important for any random testing approach nytimes.com/2020/04/24/hea…

@K_singh_P @DavidEBloom Always the case that we need consent - and raises selection problems - can be fixed statistically and usually ok if consent rate is high enough

@davidcanning6 @DavidEBloom Read your article in Boston Globe. Question: won't the sample be biased because of "consenting adult participants"? It's likely that those who were infected - or thought they were - would be more willing to participate in the study.

Our paper now up on medRxiv: The association between age, COVID-19 symptoms, and social distancing behavior in the United States medrxiv.org/content/10.110…

via @NYTOpinion nytimes.com/2020/04/24/opi…

Germany gets random testing going with sample of 3000 nytimes.com/2020/04/18/wor…

The association between age, COVID-19 symptoms, and social distancing behavior in the United States My new paper looking at how age, and having symptoms, affects people in terms of going out and close contact with others dash.harvard.edu/handle/1/42659…

How to get better COVID-19 infection data without universal testing bostonglobe.com/2020/04/15/opi…

Still the case that we lack good data Tracking site suggests White House model is overestimating coronavirus hospitalizations justthenews.com/politics-polic…

Our Harvard coronavirus survey for the US is up an running - if you are in the US please take part - we need the info to build better models and make better policies! hsph.harvard.edu/pgda/

My fried Alex posted this - really nice and smoothing - as an alternative ot our visits to the BSO facebook.com/ahealy0/videos…

There should be an end game plan for the lock down. We need to use lock downs to buy time to get infections down and put in place testing and tracking to move to a containment approach. cnn.com/2020/04/02/opi…

We need more data on COVID-19 to make better decisions around the world - I encourage you to take this survey so we can have more data! covid-19data.org

Similar to our idea for New York but for all of EU European sample tests to uncover the true spread of the coronavirus | VOX, CEPR Policy Portal voxeu.org/article/standa…

How widespread is coronavirus? We need a random sample to understand epidemic and the right policies. David Canning, David Bloom, VOX, 29 March 2020 voxeu.org/article/how-wi…

My take on "Deaths of Despair" by Case and Deaton Stronger safety nets and health care reform could help mitigate “deaths of despair”, blogs.sciencemag.org/books/2020/03/… Science 27 Mar 2020

Academic advice: seminars - when a senior professor asks a question you don’t understand always remember he is telling the audience about his work not asking about yours