One of the things I’ve been battling is the use of statistics to legitimize the stupid.
People who live in FB Land like easy-to-understand and read charts, right? The Ohio COVID-19 Dashboard has lots of interactive graphics on the page (kudos to Ohio’s IT department) but I’ve seen them being used to promote much stupid. For example, the bar graphs that show up by default are the daily count:
and it looks like everything is going down and stopping! because they don’t read the top where it says the shaded part is preliminary, i.e., they don’t have any numbers yet.
The daily chart shows exactly what the people out in FB Land should see, that we flattened the curve by doing the right things, then we opened back up a bit, and then opened pretty much everything, and even with “safety precautions” supposedly in place, we spiked. Now that masks are mandatory, it’s gone down a bit. But the chart should be cut off before that “preliminary” part to show Fox news viewers and FB users, because they’re running with big red circles around that grey bar and cutting off the “preliminary” caption, and claiming hoaxes and cover-ups.
CDC officials referred all questions to the agency’s parent organization, the U.S. Department of Health and Human Services in Washington, D.C. That suggests that HHS ordered the change, not CDC, said Jennifer Nuzzo, a Johns Hopkins University public health researcher.
On thing that really irritates me is when I see reports that compare daily covid death numbers. They’ll say “Covid deaths are down today…”
Well, you zoom in on the graph and see a startling trend emerge. The daily graph goes up and down… less people die on weekends! It’s a startling scientific find! Until you think about it and realize it’s not the virus that gets weekends off.
I suspect a big part of this is keeping schools open. COVID gets into a school, a lot of the kids who catch it will be asymptomatic, so let’s just ignore it.
while it does look like things are going down or flattening there’s a couple of odd things.
the new tests, cases, hospitalizations, and deaths all peak right at the end of july. and that’s seems really odd. there should be some delay between cases, then hospitalizations, then deaths… unless maybe a good chunk of the tests are happening at the hospitals for critical cases or something.
making things even harder to read is that ohio is still having a 5-7 day delay in test results, which none of the trackers seem to gauge. ( and actually, maybe that delay explains the hospitalizations and deaths a bit. people are getting tested, and not finding out the results until around the time they have no choice but go into the hospital. )
I would accept this explanation, which to my mind suggests that they have found manipulating data is tougher than they thought, and have settled on just not generating the data in the first place. Much easier.
They announced at some point that they were going to be adding the death numbers to the statistics on the accurate day listed on the death certificate, too. Given it can take several days for the medical examiner just to report the death, I suppose one should watch and see if the death numbers rise a week or two ago, back where no one is paying as much attention.
Over half of the police deaths so far this year have been from Covid - 19. Perhaps if they would focus on the virus and not killing POC, they’d be better off.
Good, but this desperate push to “get back to normal as soon as possible!” deeply skews the decision-making process. They should be thinking, as we are, “probably not this year, maybe not next” and committing to plans on a semester by semester basis. And clearly, in this case, the Pima County Health officials had to step in (after all, these reckless and irresponsible school reopenings basically collect COVID from around the country and then inject it into the communities where they are based).
The reason they don’t is not some pious concern over their students’ critical-academic-path-if-you-don’t-take-English-Comp-this-semester-you-might-as-well-kill-yourself bullshit. It’s about the money. Period.
With Pima, Maricopa and Pinal County expected to hit the benchmarks indicating “moderate” spread of the coronavirus this week, Harkins Theaters plans to reopen all of their Arizona locations with new safety and sanitation protocols this Friday, Aug. 28. For their celebratory opening weekend, Harkins will be screening new films like Marvel’s The New Mutants and Unhinged with Russell Crowe, as well as a special 10th anniversary screening of Infection.
Second update 8/26/2020 3:00pm: In a press briefing Wednesday afternoon, Admiral Brett Giroir—Assistant Secretary for Health at the Department of Health and Human Services and lead for COVID-19 diagnostic testing efforts—emphatically defended the changes to the CDC’s testing recommendation, saying that it came from public health experts at the CDC and was evidence based. There was “no direction” from the president, vice president, or other top Trump Administration officials, he said.
As for the changes themselves, Adm. Giroir said the decision to not recommend testing for COVID-19 exposed individuals without symptoms was intended to avoid testing too early after an exposure. This could provide a negative result before an infection has had enough time develop and register as positive on a test, thus giving an exposed person a false-assurance of being uninfected.
It’s still unclear why the CDC did not instead provide a recommended time-frame for asymptomatic testing after an exposure, particularly given that some infected people may never develop symptoms. A positive test result is necessary to ensure COVID-19 patients receive proper care, isolation instructions, and appropriate follow-up. Identifying patients through testing is also critical for contact tracing. After a person tests positive, contact tracers can inform people who may have been exposed to the COVID-19 positive person before they tested positive and/or went into quarantine.
Did I miss some research on the secondary attack rate and incubation time?
What the actual fuck is going on there?
Just FTR, Germany is just discussing easing the testing requirements because it became clear that mass testing on the scale of the last couple of weeks can’t continue due to testing capacity and availability of reagents.
no. that happened mid-july, and there’s no evidence it has changed the numbers:
… researchers at the Covid Tracking Project “explicitly don’t think the cases flattening has anything to do with the HHS changeover,” Alexis Madrigal, who helped build the tracker, told Salon.
Dr. Tom Frieden, the former head of the CDC under Barack Obama, agreed that the switch is not related to the decline in new cases. “We do have concerns about the completeness and timeliness of data being collected and reported. However, the data seem to bear out that in at least some places [there] is an actual decrease — although with still very high rates — and that is not an artifact of testing or data reporting.”
most of the data, so far as i understand, doesn’t come from federal sources. it should, but that is one of the many things this administration has left undone, blocked, or dismantled.
here’s the list – it’s mostly state and county level reporting:
while i’ve never heard that, i’m definitely no health expert. it’s entirely possible that in some places it does. if you find the bit you saw to pass along, that’d be interesting info.
