Thursday, April 30, 2020

Canada's early COVID-19 cases came from the U.S. not China, provincial data shows

K, couple things:

National Post derived these numbers by themselves. That is the first big, glaring red-flag that all but shuts down their entire premise on its own. These are journalists, not statisticians or doctors. They are not epidemiologists. They are not mathematicians.

They took two points of data: "number infected vs where those people traveled to", and graphed them, then declared the biggest destination as the source of the "early" cases. That's it. Makes zero sense, but they go ahead and leap to "thus, accurate".

Now these numbers are not the results of any sort of scientific study, and the numbers would indicate the result they are describing even if it wasn't the case. All they did was "request some numbers" (from who? article does not say). The article even points out ahead of time, "Canada didn't close off travel from China early enough". The article doesn't take into account non-confirmed cases, and specifically cuts off its date range at Jan 15th. Weird place to cut it off, imo: That's after China began locking down their country. Which is why you don't see China in the lists cited by the article at all.

Basically, they're looking at confirmed cases and comparing "where you traveled to recently" then leaping to "...so that's how they got it". And no matter how they got the virus, in Canada, the clearest and biggest majority of international travel is and always has been to and from the US. If we remove the #1 spot, we're left with the UK and France. Does the article try to make that same connection -- "they brought it back from France or the UK"? Of course not. Because that's besides the point. The point is "America bad".

Because just the assumption that "travel == caught the virus" is bunk as hell. Canada's entire infected population could've come from one single person on a cruise ship and these numbers wouldn't look any different.

So take this all with a big fat grain of salt. No statistician or epidemiologist would put their name on work this shoddy.

No comments:

Post a Comment