Tai sitten raportointi pettää. En ole numeroita tarkastanut mistään muusta lähteestä joten voi olla puhdasta foliohattuilua, mutta 16.7. datan käsittelijä vaihtui CDC:stä HHS:ksi (Department of Health & Human Services), jonka kirjoittaja väittää politisoituneen pahasti.
Kommenteissa on myös muita käyriä, ja myös seuraava kommentti alkuperäiseltä kirjoittajalta:
I'll leave most of this without comment, other than the fact that I do not have as much visibility into the reporting structure as I wish I did. What I bring to the conversation that is new, I believe, is a serious, detailed look at the data being reported.
Among the facts that any explanation must address:
* A break occurred that was strong enough to take close to the most rapidly rising case count in the world and abruptly make it flat. Things that massive and dispersed with that kind of momentum don’t turn on a dime.
* When all data are combined, that break was as discontinuous as it is possible for such data to be.
* That break occurred on the day responsibility for the data was taken away from a scientific organization and handed to a manifestly political organization.
* The sudden change affected only states that are politically aligned with the President.
* Analysis of those data show that the statistical properties of red states data diverged dramatically from the statistical properties of other data at that moment.
* Suggestions such as “HHS pulls the same data at a different rate” predict that the effects show show up equally in all state data. They do not. They only show up in red state data.
* Handling of the data was taken away from CDC and given to HHS. That was clearly not a statistical fluke. LOL. That intentional action was intended to serve some purpose. The President said pretty plainly that his reason for taking the data away from the CDC and handing it to HHS was because he didn’t like the numbers coming out of the CDC. I am inclined to take him at his word, which leaves me unsurprised.
Is it possible that all of that is just coincidence? I suppose. But alternative ideas have to be looked at as critically as my conclusions. “It’s all a big statistical fluke that just happens to align with Trump’s political agenda” doesn’t really pass a smell test.
