When I was 8 or 9 years old, I did a "reading comprehension" test about a reading that involved a child picking up a scorpion by the tail. I got the question wrong b/c apparently I was supposed to realize that "by the tail" was the most dangerous way to pick up the scorpion.
I thought it was self-evident that when picking up a scorpion you would absolutely want to hold it by the tail.
This says something about the childhood I had had to that point, compared to the person writing that standardized test.
I also struggled on a placement test with putting pictures in order showing the steps to order a pizza delivery. The part of Africa we had just returned from did not have pizza delivery.
Amusing stories...
until you realize that these tests are used to decide what opportunities we have as we grow up, and then think of the impact on people from different backgrounds.
• • •
Missing some Tweet in this thread? You can try to
force a refresh
A comment on whether the "Doherty Model" is appropriate if case counts are high when threshold of 70% or 80% is met:
[note - this is my personal opinion without consulting collaborators, should not be taken as official statement of any group I'm affiliated with]
These thresholds provide valuable guidance to policy makers and the public to help them develop plans, and to see that vaccine is the best tool to get us out of our current predicament.
Modelling always involves assumptions, and there is always a risk that an assumption is wrong in a way that materially affects the outcome.
In this case a key question has been raised: might the case count be so high that contact tracing and similar interventions can't keep up?
Why do lockdowns become more important just before or in the midst of a vaccine rollout?
1/n
First, let me both dispel and validate one criticism of lockdowns: "you're just delaying the infections - they will happen later"
2/n
If lockdowns or any other intervention happen but at the same time some immunity builds up, then the epidemic peak will be lower and the total number infected will be smaller.
(seasonal effects may complicate this claim).
Flattening the curve does reduce total infections
3/n
In the thread below, the claim is made that COVID-19 is only hypothetically worse than common cold and that the deaths are a result of lockdown and fear rather than COVID-19.
When one makes a hypothesis, one should look for other, simpler hypotheses which could also explain the data.
In this case, perhaps one might think that the interventions done to control COVID-19 might also control influenza. Since flu has a lower R0, this might be enough.
Another thing a scientist should do is to look for other data that might refute or support the hypothesis.
In this case, perhaps one might look at countries that didn't have a large COVID-19 epidemic. For example Australia and New Zealand.
Some comments on the VIC path to getting its epidemic under control. A 🧵
Victoria hit over 700 cases a day, and now has gotten into single digits per day.
When cases first started doubling each week, there were a few relatively mild restrictions put in. I believed at the time it was an appropriate scale of response. Many felt it was an over-response.
Cases kept growing with a similar rate. More restrictions were put in, and then finally when cases continued to grow, they put in place a fairly strict lockdown (about this time we hit 700 a day).