I suppose it would be more accurate to say that is what they are reporting today. They are reporting about half as much testing as they were two weeks ago with a still very high average positivity rate. Better?
The other dumb thing about “KIDS NEED TO BE BACK IN SCHOOL” is that it’s not like the in-school experience is going to be good. Kids and teachers will be so distracted, trying to stay safe (if things are done even halfway right), fidgeting with masks, and wondering if the tiniest cough from across the room is going to infect them.
Who knows how my kids’ remote learning is going to be, but I’ll guarantee they will be able to focus better at home (ok, maybe I shouldn’t guarantee that when they have the internet and computer games at the ready).
2 Likes
Not criticizing you, just my own belief about how to interpret positivity rates has changed recently, so I’m kind of talking my thoughts out loud.
I assume that there’s a lag of test results after testing, similar to the lag between new positives and deaths. (Presumably that lag between testing and results is only a few days.) If there’s a sharp drop in cases, what that would mean is that current test results (from lagged tests) are being reported today, along with today’s report of current tests (which are happening at a lower level).
So even if the true positivity rate were constant, the sharp drop in testing would result in an artificial increase in reported positivity rate. That’s why I don’t think we can interpret positivity rate at all in this case.
I guess I assumed even if these results were lagged that the tests/results were on the same day. Is that not right?
[I am getting flagged by the software that I am maybe participating too much in the topic - Sorry!]
That’s the question we were trying to figure out in the earlier conversation from this morning. Prior to today, without putting much thought into it, I had just assumed that:
- Daily reported tests were the number of tests administered that day, and would be a good measure of testing activity.
- Daily positivity was the number of new cases in those tests, and would also be a good measure of whether testing activity was sufficient and if the virus were spreading in a community
But after thinking about it for 5 minutes, there’s no way that both of those can be true, because we’re seeing widespread reports of delays between testing date and reporting dates. (Maybe it’s easier to think of this as the gap between specimen collection and specimen analysis.)
I think the daily new cases number is the one I feel most confident in. I think it’s very likely to be the number of newly reported positive cases for that day (regardless of when the test was administered).
But the daily tests number could be:
- The number of tests administered that day, regardless of when the results are known. This is my assumption, although (for reasons described earlier) it means that a calculated positivity rate based on this number and new cases is meaningless.
- The number of new test results for that day, regardless of when those tests were administered. If true, this would make the positivity rate more meaningful, but make this a not-as-good measure of states’ recent testing activity.
- Something else?
The way he accepts being carried out by his son like that makes it look like it’s not the first time.
6 Likes
Nope. Pretty much spot on.
So we have pro sports team in a bubble-looking good so far
Pro sports teams not in a bubble. LOL MLB
College athletes working out. Is 15 cases at one school too many??
College kids in genera back on campus. You’ve got to be effin kidding me.
1 Like
I would assume the labs are reporting test results to the state as they run them. So maybe it works like this:
-At some time in the past tests are taken.
-At some point 4-10 days after that the tests are processed.
-The lab reports all tests they processed that day to the state regardless of when they were collected.
I would be very surprised if there was a two stage reporting mechanism where the labs were reporting how many samples they got in each day and then later reporting the results of them. Although even under my example that still means the numbers we are getting today are from like a week ago which is a still a problem.
This seems like as good a guess as any, and it would mean that positivity rates are a meaningful statistic. But, as you noted, it means that it’s not giving the up-to-date activity that (up until this morning) I thought it was. And if there’s non-random selection that determines which specimens are tested quickly vs. not, then you can throw everything out the window.
1 Like
I’ve resolved that I’m going to do one of two things if I run into an anti masker in public like that.
#1 I’ll offer them one of my daughter’s pacifiers
Or
#2 I’ll start streaming Fox News on my phone and show it to them, all while saying, “here boy (or girl)” and trying to walk them out of the store like a they’re a dog and I have a treat.
I’m busy in something (work!). Can someone else run the formula.
Cases x (2.5 + 16 x positivity^0.5) and see how consistent the numbers turn out for Florida?
For 17 % it works out to (2.5 + 16 x (.17)^0.5)=
2.5 + 16 x 0.41= 9.1 roughly.
So 5,000 cases x 9.1 is 45,500 actual cases.
Maybe just take 7DRA of tests and positives and do the calc from their for smoothing? Should eliminate some of the odd reporting lags.
I like dickin with the holier than though types that throw their Christianity out the window for SPORTS. Assume I will get no reply. But it’s fun nonetheless
11 Likes
I think this is correct. The AZ dashboard summary reports a daily “number of new tests reported today” but if you drill down, you can find a graph of test results by date of collection.
So right now it’s showing counts for the past few days as:
8/7 8939
8/8 4881
8/9 2144
8/10 21
So these numbers will go up for a while and then eventually stabilize at the actual number of tests that were done on a specific day.
2 Likes
Among my more “optimistic” friends, when they’re talking about school openings, I’ll say it’s not likely schools will be able to remain open and that cases/deaths are likely to spike. When they talk about real estate, I’ll bring up evictions. Suddenly they go completely quiet. They don’t want to engage the crazy guy.
1 Like
Want the “college experience”, but your school is going remote? Try The U Experience, where you can do your online coursework while sharing a “bubble” in a hotel in Hawaii with other similarly situated students from around the country. In addition to getting to “quarantine” in Waikiki, you can participate in excursions to private islands, hikes in nearby mountains, and private parties.
Everyone has to take a vivid test before entering “the bubble”, so I’m sure this will be fine…
I have been doing three hour beach walks and I’m not gonna lie, cotton masks suck. Gaiters are better but they also stink.
I start at 5:30 now, walk on the sand not the well-trafficked (after 7) paved trail, and don’t wear a mask.
I have done the entire walk in an N95 and KN95 and frankly find them both more comfortable. Cotton gets soaked.
Lol at the sign at the entrance: SO GOOD to see you!
He was at fucking Sprouts! He had to have gone there specifically to pick this fight.
4 Likes
No one knows, so that’s why the data is such a mess. It seems quite clear a place that reports only positives on any day shows the testing was probably reported prior to the result being given.
I’ve seen enough graphs to be fairly confident testing and positives should rise and fall together and shouldn’t really have an offset. If a place does, that suggests reporting shenanigans to me. I think around ten places look off right now, but it might be less.
This is what you get when you weaponize a virus against the citizens of USA#1:
1 Like
This reminds me of something my friends with kids in daycare have reported: apparently it’s becoming normal for daycares to offer “onsite remote learning” to older kids—as old as middle school.
So basically the government says it’s not safe for kids to go back to school, so parents are stuffing them in even more crowded daycares that are haphazardly modified to babysit older kids during remote learning.
Seems fine, USA #1.