Paul Cantrell
Do we the same number of car accidents as last year as “zero?” Same number of murders?!
I can think of no earthly scientific reason for this. This is data propaganda. Change my mind.
From @violetblue: https://mastodon.social/@violetblue/113852273879261711
Darrin West
@inthehands @violetblue Agreed. Normalizing the data to zero means they are saying this level is “normal”, and expected now. Evil.
Paul Cantrell
@violetblue
OK, so…no, they’re not using last year’s total or average as a baseline; they’re taking the 10th percentile of a 6-month period. That’s a perfectly reasonable metric for “about as low as it’s been recently.”
That part of the quoted post was wrong.
Details here: https://www.cdc.gov/nwss/about-data.html
The part that wasn’t completely wrong…
Paul Cantrell
…is that they •are• reporting COVID activity levels relative to that baseline.
What that means is if COVID levels are about as low as they’ve been recently, the CDC will use the official word “minimal” for that.
Whether “minimal” gets treated as “zero” in the public mind…well, that depends on the press and the Trump administration doing a good job. 🙄🙄
Paul Cantrell
@waldoj Yeah, see updates in thread following the OP.
Jamie McCarthy
@inthehands @violetblue That’s not what the CDC says. That’s what someone who didn’t read assumed.
Go see for yourself on the CDC page: see “About the Data” and “Data Methods.” The baseline is the 10th percentile of previous timeframes. Which seems reasonable to me.
Not everything you read online is true.
Waldo Jaquith
@jamiemccarthy @inthehands @violetblue Yeah, I'm reading through this and just don't see anything like what's claimed in the original post. This description is pretty clear, and it is definitely not “last year's average is this year's zero.”
Paul Cantrell
@nazokiyoubinbou I added more details in a thread from the OP that you may find helpful.
Nazo
@jamiemccarthy @inthehands @violetblue I'm not really clear on what this means, but I can't help but to notice in the screenshot it saying "for site and method combinations (as listed above) with over six months of data, baselines are recalculated every six calendar months (January 1st and July 1st) using the past 12 months of data."
That seems to be saying that the baseline will indeed be calculated comparing to the previous year.
I apologize if I'm misunderstanding. Can you explain better?
Jamie McCarthy
@nazokiyoubinbou @inthehands @violetblue The key part is: the baseline is the 10th percentile of the past 12 months of data.
When comparing to the previous year, using the 10th percentile is a good approximation to the lowest amount that's reasonable to find. Especially for data that's as flaky as wastewater. It's hard to measure viral load in wastewater.
So, this year's baseline is last year's 10th percentile. Seems reasonable.
Nazo
@jamiemccarthy @inthehands @violetblue Again, I really don't understand what that means, hence asking. But even trying to take this as meaning something other than it being based +/- on last year's levels it still sounds like something dangerously at risk of minimizing the levels in people's minds.
Paul Cantrell
In short:
What CDC is doing is •not• like saying “we are going to call the same number of murders •zero•.”
What the CDC is doing is saying “we are going to call the lowest the murder rate has recently been •minimal•.”
That’s…not awesome, but not as egregious as the OP made it out to be.
Paul Cantrell
@jamiemccarthy
Thanks for the heads up. I dug up the text myself (link for anybody reading so they don't have to dig too: https://www.cdc.gov/nwss/about-data.html#), and the OP definitely gets it wrong.
I added an explanatory thread after the OP.
Paul Cantrell
Thinking about my own hotheaded reaction to the OP, and the dread I feel about the likely social impact of the CDC’s not unreasonable but kind of weak sauce policy, my thoughts keep coming back to how hard it is to live in a low-trust environment.
If I trusted the CDC to communicate well and be forceful in pursuit of public health…
If I trusted the press to communicate this to the public well…
If I trusted the incoming administration not to twist every weakly stated truth into an outright lie…
…but yeah, I don't. None of the above.
Paul Cantrell
I will give a shout-out to the hero of this conversation, though: this community.
Someone in the replies respectfully pointed out that I was spreading bad info.
Mastodon post editing let me notify everyone who reposted or favorited the OP that there was a correction.
And shout out to the OP for raising this issue to my attention in the first place. That’s important too. We can’t get everything perfect on the first try in our social media posts, but that shouldn't stop us from communicating! I wouldn’t have learned any of this if it weren’t for them.
GNU social JP is a social network, courtesy of GNU social JP管理人. It runs on GNU social, version 2.0.2-dev, available under the GNU Affero General Public License.
All GNU social JP content and data are available under the Creative Commons Attribution 3.0 license.