Pingui posted:

There is this weird thing going on, where basic logic is being cast aside, because it hasn't been ~scientifically proven~.

Pingui posted:

There is this weird thing going on, where basic logic is being cast aside, because it hasn't been ~scientifically proven~. Idk if it is just the general alienation, if it is politics-brain, increased complexity and esotericity of research or some offshoot of Freakonomics, but its been a battle throughout the pandemic. I wish it was possible to incorporate presumptive defaults into statistics, because too often studies are simply underpowered to a statistical significant degree, which is invariably noted as "no effect", rather than a likely effect that should be studied more rigorously and in depth.

As a recent example see the Paxlovid study for non-hospitalized veterans post 30 days, we know COVID fucks with cardiovascular health for this specific group and even at the tail-end we can see the positive effects, but the confidence intervals are too great and so the takeaway becomes "it doesn't work". But the biological mechanisms dictates that it works, the data shows a decrease on average even if the confidence intervals extend slightly across no effect and early studies demonstrate that they are looking at the tail-end of the event distribution - in aggregate it logically follows that it loving works.

It is immensely frustrating and completely detrimental to a scientific process imho.

Pingui posted:

There is this weird thing going on, where basic logic is being cast aside, because it hasn't been ~scientifically proven~. Idk if it is just the general alienation, if it is politics-brain, increased complexity and esotericity of research or some offshoot of Freakonomics, but its been a battle throughout the pandemic. I wish it was possible to incorporate presumptive defaults into statistics, because too often studies are simply underpowered to a statistical significant degree, which is invariably noted as "no effect", rather than a likely effect that should be studied more rigorously and in depth.

As a recent example see the Paxlovid study for non-hospitalized veterans post 30 days, we know COVID fucks with cardiovascular health for this specific group and even at the tail-end we can see the positive effects, but the confidence intervals are too great and so the takeaway becomes "it doesn't work". But the biological mechanisms dictates that it works, the data shows a decrease on average even if the confidence intervals extend slightly across no effect and early studies demonstrate that they are looking at the tail-end of the event distribution - in aggregate it logically follows that it loving works.

It is immensely frustrating and completely detrimental to a scientific process imho.

Pingui posted:

There is this weird thing going on, where basic logic is being cast aside, because it hasn't been ~scientifically proven~. Idk if it is just the general alienation, if it is politics-brain, increased complexity and esotericity of research or some offshoot of Freakonomics, but its been a battle throughout the pandemic. I wish it was possible to incorporate presumptive defaults into statistics, because too often studies are simply underpowered to a statistical significant degree, which is invariably noted as "no effect", rather than a likely effect that should be studied more rigorously and in depth.

As a recent example see the Paxlovid study for non-hospitalized veterans post 30 days, we know COVID fucks with cardiovascular health for this specific group and even at the tail-end we can see the positive effects, but the confidence intervals are too great and so the takeaway becomes "it doesn't work". But the biological mechanisms dictates that it works, the data shows a decrease on average even if the confidence intervals extend slightly across no effect and early studies demonstrate that they are looking at the tail-end of the event distribution - in aggregate it logically follows that it loving works.

It is immensely frustrating and completely detrimental to a scientific process imho.

Crazyweasel posted:

Son has had runny nose, cough, low grade fever for 2 days.

Suddenly last night I get this weird metallic taste in my mouth I can’t shake but eventually it kinda goes away and I go to sleep. My wife says nothing has changed for her being in the same room and all, so don’t think anything environmental.

Today my morning coffee has a long bitter aftertaste, but I’m on high alert and low sleep so maybe I just made a lovely cup.

My COVID tests with May expiration for me and the boy negative. I’m sure even if it is COVID we’ll probably never test positive anyways.

SardonicTyrant posted:

It's really straightforward, honestly; the government doesn't want to protect people from covid because that is expensive and gives labor more power, so they won't help people; and people fundamentally can't maintain a lockdown on an individual basis so they gravitate to the "factoids" that say covid is over and you can open biden as long as you're "careful". People don't act "rationally", that's a fake thing. People as a whole take the path of least resistance. And if the path of least resistance leads to disability and death....well, welcome to the wheel of samsara.

Real Mean Queen posted:

I think it's just where we're headed from now on. You see it everywhere now, there's this very "well you'd better have a study backing up every word in that sentence, and I'd better like them" approach to everything, because the wrong people figured out that this is an easy way to not get things done. It works, too

Snowglobe of Doom posted:

The part where everyone outside this thread (and certain pockets of reddit and twitter) all agreed that covid doesn't spread in schools was what broke my brain

The part where they all agreed to wear masks inside cafes and restaurants except when they were sitting and eating was also , it's just the most blatant motivated reasoning. If they could at least have been open about "This will increase spread somewhat (which will also increase hospitalizations and deaths) but we see that as an acceptable tradeoff if parents can keep going to work/restaurants can stay in business" that would have been a different matter, but of course they couldn't just come out and say "We're allowing more deaths so $$$$$ can keep circulating" so everyone had to clap and believe in fairies to make the magic happen

Pingui posted:

Yup and it was a stellar example of motivated reasoning and perhaps that is really a big part of this. Not that the researchers are necessarily applying motivated reasoning (though, lol, it happens a lot), but because the foundation of motivated reasoning results in the mentioned self-censoring. Like there is a lack of rationality at an emotional level, so when the data shows an ugly truth it gets worked against, the data massaged a bit, conclusions become very stringent and narrow, the discussion lackluster and uninspired etc.

Pingui posted:

There is this weird thing going on, where basic logic is being cast aside, because it hasn't been ~scientifically proven~. Idk if it is just the general alienation, if it is politics-brain, increased complexity and esotericity of research or some offshoot of Freakonomics, but its been a battle throughout the pandemic. I wish it was possible to incorporate presumptive defaults into statistics, because too often studies are simply underpowered to a statistical significant degree, which is invariably noted as "no effect", rather than a likely effect that should be studied more rigorously and in depth.

As a recent example see the Paxlovid study for non-hospitalized veterans post 30 days, we know COVID fucks with cardiovascular health for this specific group and even at the tail-end we can see the positive effects, but the confidence intervals are too great and so the takeaway becomes "it doesn't work". But the biological mechanisms dictates that it works, the data shows a decrease on average even if the confidence intervals extend slightly across no effect and early studies demonstrate that they are looking at the tail-end of the event distribution - in aggregate it logically follows that it loving works.

It is immensely frustrating and completely detrimental to a scientific process imho.

Pingui posted:

"Effectiveness of Nirmatrelvir–Ritonavir Against the Development of Post–COVID-19 Conditions Among U.S. Veterans"
...

gradenko_2000 posted:

Haven't posted here in a while but new not-covid just dropped

https://twitter.com/PhilippineStar/status/1720574229173997732?t=HG-HH7flIFZqDpdi4hod6Q&s=19

Pingui posted:

PerniciousKnid posted:

Edit: also dude was shaky as hell and stabbed my kid an extra time so there was blood everywhere.

PerniciousKnid posted:

It's impossible to get kids under 6 flu shots outside of school hours unless you pay $25 at urgent care. But if I keep the kids home it's illegal truancy.

FuturePerfect VideoGames posted:

I went back and looked at that and thought it was notable that the only cardiac symptom that was pax was definitively "not helpful" for (after 30 days) was chest pain. Four others were on the (perhaps marginally) effective side (I suppose dysrhythmias and myo/pericarditis might be truly no difference):
(..)

FuturePerfect VideoGames posted:

(..)
... but the confidence internal for total cardiac seems rather large? I can't quite eyeball if they just added everything in quadrature correctly but it also seems notable that the cardiac cumulative incidence rate is much smaller than the sum of the subcategories, implying that (presumably) many people had 2 or more symptoms. I wonder if they'd measured "average # of cardiac symptoms" or some kind of severity-weighted index, if that would have been more convincingly in favour of pax.

https://www.acpjournals.org/doi/suppl/10.7326/M23-1394/suppl_file/M23-1394_Supplement.pdf posted:

(..)
*For each post-COVID condition, 31-90 day incidence is calculated only in matched groups within which all persons do not have prevalence of the condition of interest at baseline (i.e. not documentation within 12 months prior to infection).
(..)

Pingui posted:

There is this weird thing going on, where basic logic is being cast aside, because it hasn't been ~scientifically proven~. Idk if it is just the general alienation, if it is politics-brain, increased complexity and esotericity of research or some offshoot of Freakonomics, but its been a battle throughout the pandemic. I wish it was possible to incorporate presumptive defaults into statistics, because too often studies are simply underpowered to a statistical significant degree, which is invariably noted as "no effect", rather than a likely effect that should be studied more rigorously and in depth.

As a recent example see the Paxlovid study for non-hospitalized veterans post 30 days, we know COVID fucks with cardiovascular health for this specific group and even at the tail-end we can see the positive effects, but the confidence intervals are too great and so the takeaway becomes "it doesn't work". But the biological mechanisms dictates that it works, the data shows a decrease on average even if the confidence intervals extend slightly across no effect and early studies demonstrate that they are looking at the tail-end of the event distribution - in aggregate it logically follows that it loving works.

It is immensely frustrating and completely detrimental to a scientific process imho.

Thoguh posted:

Sitting on a beach in Mexico right now. It’s a spreadsheet avoidance vacation but honestly it wasn’t super hard to find a place that is almost completely outside, including the restaurants.

Raskolnikov2089 posted:

anus (butthole)

The Demilich posted:

Woke up tasting metal the other day, but I figured it was because I was scraping my nasal cavity with the little test qtip (tested negative btw)

Real Mean Queen posted:

I think it's just where we're headed from now on. You see it everywhere now, there's this very "well you'd better have a study backing up every word in that sentence, and I'd better like them" approach to everything, because the wrong people figured out that this is an easy way to not get things done. It works, too

Myok posted:

I told my mom about this and she promptly went and scheduled her booster.

Why Am I So Tired posted:

Jealous. My wife's dad has been sick for 3 weeks now and his wife for 4 with "not COVID" (they took one test 3 weeks ago, you see), they both rebounded and everything. My wife told her mom about it as a warning, and you'd think that would scare her, but yesterday she dropped her "I always wear a mask" charade or decided now was a good time to finally drop masking for some reason and FaceTime'd us maskless from a vet's waiting room and wouldn't put one on when we begged. I really wish people weren't like this!

no lube so what posted:

hits home. sorry. that's so frustrating.

some of my family has been visiting sick friends in the hospital because of a new autoimmune disorder destroying there liver. I gave the fam a box of aurora & a box of vflex. declined, they live in a new paradigm. sigh.

Thoguh posted:

completely outside, including the restaurants.

Steve Yun posted:

https://twitter.com/inkblue01/status/1720785263134187749?s=46

https://twitter.com/moderndarkage/status/1720821758599573944?s=46

Thoguh posted:

I flew yesterday and there were almost zero masks other than my wife and I. Maybe 1-2% of people on the flights. When I flew to Alaska a few months ago masks were probably closer to 5% of passengers. Not common but not rare. This time we stuck out as the only people.

Sitting on a beach in Mexico right now. It’s a spreadsheet avoidance vacation but honestly it wasn’t super hard to find a place that is almost completely outside, including the restaurants.

gently caress COREY PERRY posted:

Shiroc posted:

Gotta admit I'm pretty bummed about there not actually being anything on the horizon that will change the incredibly stupid situation we're in.

Pingui posted:

Which makes sense within the confines of communication to the public, but what I am talking about is more insidious and pervasive than that. There is a bunch of research that seems honest, but self-censoring to only let the statistics do the talking and the rest of the paper is just packaging. It doesn't seem inherently malicious, but it makes for some very poor science.

So while I agree that what you are describing happens, it isn't all decision-based evidence-making. At least I don't think that explains things like the Paxlovid study.

Yeah, that is what I meant by politics-brain; like the data, research and conclusions are inherently written as political in the US sense. Good vs evil. Right vs wrong. Us vs them. It is certainly an element when talking COVID research and tbf COVID research is difficult to dislodge from politics. I can imagine the effect is to play it defensively and only let the statistics do the talking, which is why I mentioned it.

But like above, while I think you are correct that this happens, I think that more relates to the communications in media and doesn't (to me) explain research like the Paxlovid study I mentioned.

Yup and it was a stellar example of motivated reasoning and perhaps that is really a big part of this. Not that the researchers are necessarily applying motivated reasoning (though, lol, it happens a lot), but because the foundation of motivated reasoning results in the mentioned self-censoring. Like there is a lack of rationality at an emotional level, so when the data shows an ugly truth it gets worked against, the data massaged a bit, conclusions become very stringent and narrow, the discussion lackluster and uninspired etc.

Raskolnikov2089 posted:

[Pestilence] %2C%20or-,anus%20(butthole)

captainbananas posted:

I dunno if you play the peer review publishing roulette game yourself, Pingui, but I immediately saw what you saw and thought: this is exactly the kind of poo poo anonymous reviewers and/or editors will latch onto and fight authors about. A lot of them cannot deal with having contextual language discussing how one or more null results are *not* in fact the same thing as saying there is definitely no effect, but that the particular analysis might be underpowered, not purposively calibrated for the particular effect, whatever the case may be. If reviewers or editors want to play hardball about it it can add months to the time from submission to publication. So yeah, when you say that some of it seems "self-censoring," that could either be censoring from the peer review process or from the author having been conditioned by past experiences.

I have dealt with this in my own work and I don't work on anything half again as politicized as SARS2. Though I did contribute to a public health messaging experiment project that was sabotaged by a right-wing coauthor at the 11th hour to avoid making conservatives look bad. That was loving magical. gently caress you, author #2.