The White House has been turned into slave quarters for its teachers’ union masters, and therefore must do whatever those unions want, and in turn the CDC is ordered to produce (i.e., make up) research findings to justify the policies the teachers’ unions want, regardless of what the data says, so the agency does its desperate best to sate the insatiable lust of those unions for masking every child for the rest of their lives. One garbage study after another, with the most recent coming at the end of this week. Here is the press release announcing studies supporting masking in school, with links to each paper. (CDC Garbage)
One of the studies, a supposed comparison of CV-19 rates among children in counties with and without a school mask requirement, is an absolute embarassment that is the worst piece of dreck yet, so bad that it doesn’t even use statistics properly. Most people don’t know what a T test statistic does, and probably shouldn’t be expected to, but the idiots at the CDC are supposed to know how to analyze data and they clearly don’t how to use, or when to use a T test, which is what they applied to this data analysis. (If you want a basic explanation of the T test, see this, and after reading you will apparently know more than the CDC “experts” do. (T test)) A one-sided T test was used, which means they assumed there would be an effect. I thought the purpose of research was to ascertain relationships, not presume them. Why use this statistical approach–because it is the only way to get the result that you want.
Now what is the first thing that we look for in a CDC study? That’s right, a cherrypicked time period. Now what is the second thing we look for in a CDC study? Correct again, completely ignoring important and relevant potential confounders. The time period here is very short, just a few weeks, why wasn’t last year included? Why don’t we look at the longer trend in pediatric cases and see if there were changes in that trend. Note that although the study is looking at school mask policies, the cases included are all cases in children, including those who aren’t going to school. And the unit of measurement isn’t a school, it is a county, so counties which had some schools with mask requirements and some without were excluded. Wonder what the effect of that exclusion might be–oh, see the study below. And, as always with these masking studies, no assessment of where and how transmission actually occurred. Just a made-up inference that it must have been associated with school.
The other study purported to show that there fewer cases in schools in two large counties in Arizona that mandated masks versus schools that did not, at the start of the school year. Let’s start with the basics. The study isn’t looking at cases, it looks at “outbreaks”, which is two or more cases at a school. So no assessment of whether the cases were transmitted in the school, in other words, the study tells you absolutely nothing about actual transmission in schools. The major statistical trick here is that they tried to eliminate schools which had a mask mandate enacted after the school year started from the analysis. Why? Because if you include them 52% of schools with a mask mandate had an outbreak versus 48% of schools that didn’t. If you leave out the “late” mask starts, you can claim that only 21% of schools with a mask mandate had an outbreak. Oh, and did they indicate when the outbreak occurred in the schools with the “late” mask mandate. Of course not, even though the data was available to them, because, again, that would screw up the message they were ordered to deliver. You have to assume, based on the way these jackasses do things, that if you looked at the actual date of the outbreak compared to the when the mask requirement went into effect, you will find absolutely no difference between the schools.
I don’t know why I bother reviewing these CDC studies when other people do a fantastic job. Here is a beautiful explanation of how bad these studies are. (Prasad Review) In a normal world that actually believed in science, they would never pass peer review, in fact they would be laughed at.
If you want a little more sane perspective on masking, read this review of the research. (Masking Article)
Join the discussion 6 Comments
Thanks much for this detailed explanation.
The twin cities marathon is sure buying into their garbage. Runners are being forced to literally wear a mask until they cross the start line and put it back on at the finish line. No word in whether there’s an award for completing a marathon with an actual Covid infection 😁
Thank you for all of your hard work Kevin. I check your site every day to see the latest nonsense being foisted upon us.
I’ve looked at all of this mask data for my site and found it similarly wanting. But I try on my site not to make any guesses as to motivation, whereas it doesn’t appear that you have such reservations.
So I’ll pose this question to you: as Adam Carolla says, “Stupid or liar?” Are the people publishing all of these studies this incompetent that they don’t even know the basic errors they’re making? Or are they well aware of the errors, but know that most people won’t look carefully enough to find them, or will be blinded by political tribalism?
I suspect they know what the message is supposed to be and at a minimum are implicitly aiming for results that justify the method. But it is hard not to thing it is intentional when they make some of the design and statistical analysis choices they make. I doubt they are really incompetent.
Josh Hamer- you mention Carolla, and he’s made an interesting observation over the past week, namely that the blue collar types he knows are generally non panickers- his explanation is that the people who work for a living are accustomed to messiness, dirt, and the fact that stuff happens- while all his Hollywood pals are in a fetal position hiding under the bed. I see that in my world, substituting academic types for Hollywood pukes. NBA stars and stand up comics are making more sense than public health officials.
Nothing leaves the CDC for public consumption unless it’s reviewed by several layers of liars, especially under today’s circumstances. To even hint at ‘stupidity’ or ‘incompetency’ as an excuse for this ‘Garbage’ is insulting to us all. To Kevin’s point, they start with the narrative as the goal and spend countless hours of resources from many layers of the organization to craft the data to match. There’s NOTHING accidental leaving this polluted agency.