In Case You Thought I Was Kidding

 

I wasn't kidding. I'm just getting warmed up. Part One continues...

1) Was This Trip Really Necessary?

We've shown the clowncar of earnest idiots in the economics department at Johns-Hopkins set out on a fool's errand in the first place with this "study". But as they open, they explain why they decided to do it anyways.

"Our early interest in the subject was spurred by two studies. First, Atkeson et al. (2020) showed that “across all countries and U.S. states that we study, the growth rates of daily deaths from COVID-19 fell from a wide range of initially high levels to levels close to zero within 20-30 days after each region experienced 25 cumulative deaths.”

So, some idiot suggested COVID deaths would level off after 25 deaths. (SRSLY??) And two years and millions of dead people later, this gang of idiots thought they needed to see if this was, in fact, the case. 

To be fair, this is actual scientific inquiry (hypothesis + experimentation = conclusion), but it's actual science from, like, retarded fetal alcohol syndrome kids fed lead paint chips and drinking out of lead mugs. It's on the same level as Dr. Robert Ballard writing a grant proposal to study whether the RMS Titanic was actually unsinkable. After he discovered the wrecked pieces of it on the ocean floor 73 years later. "Chief, I'm thinking your study is a bit late on locking the barn door after the horse already got out there."

We'll get to some actual numbers in a minute, but they really open the ball with the following sentence:

"Today, it remains an open question as to whether lockdowns have had a large, significant effect on COVID-19 mortality."

SRSLY?!? In what psych ward???

 Walk with me for thirty seconds. If you had looked at nearly three dozen studies of different people lowering the temperature of water until it froze, and controlling for the other variables (atmospheric pressure, salinity, etc.) Who The Actual Fuck would open their "study" by saying

"Today, it remains an open question whether water freezes at 32° Farenheit."???

If you had that many relevant, properly designed studies, the answer is "No one with an IQ above fungus would ask that question." In fact, to even ask the question, let alone declare it "open", you'd have to be a functional retard. Or, a Johns-Hopkins economics researcher. But I repeat myself.

The question being "open" implies that either the studies point in all directions, that they must not be repeatable, accurate, nor well-constructed or valid, or that the people looking at them cannot comprehend what they're reading.

Let's drill a little deeper on how they worked this out.

2) Meta-What?

This "study" in question is no such thing. No actual research was done. Because it's a meta-survey. Think of it as the Cliff notes to multiple other studies, purporting to have any internal errors in any one study cancel each other out, to better reveal the truth.

It's also the lazy man's way to sound smart without doing any actual, y'know, science.

 

"A meta-analysis is a statistical analysis that combines the results of multiple scientific studies. Meta-analyses can be performed when there are multiple scientific studies addressing the same question, with each individual study reporting measurements that are expected to have some degree of error. The aim then is to use approaches from statistics to derive a pooled estimate closest to the unknown common truth based on how this error is perceived. Meta-analytic results are considered the most trustworthy source of evidence by the evidence-based medicine literature." - Wikipedia: meta-analysis

Sounds positively spiffy, and it can be, with a few caveats.

The surveys or studies considered have to be looking at the same question, the same way, and controlling out all the same variables. Otherwise you're just playing Splattergories.

They should all be properly constructed, with valid methodology.

The results for each should be quantifiable, not subjective, and in a common frame of reference, not apples-to-oranges, let alone houseflies-to-houses.

Good ones are always this, which is what makes them useful in the medical community. Bad ones are none of this, which makes them towering piles of bullshit masquerading as science. 

For example:

Suppose you have in your employ one Faston Flock, the great grand-nephew of Himself, John Moses Browning, and he has invented, designed, and built a truly splendiferous and earth-shattering method of making accurate rifles, and you want to see how accurate they really can be.

So you assign ten product reps, one at a time, to take three rifles, and 39 rounds of ammunition, to each of the top ten rifle shooters your research department can find. They're all top marksmanship competitors, based on points in international matches, Olympic events, pre-qualifying matches, etc. They're all right-handed and right-eye dominant (to eliminate another pair of variables). They each get 13 rounds of ammunition in three lots named Red, Blue, and Green. They get three spotters for each lot, and after sighting in, they fire ten rounds for grouping, from three rifles, #1, #2, and #3. Rifle #1 is fired with Red, #2 is fired with Blue, and #3 is fired with Green. No time limit is allotted. The reps each record the time of day, the weather (temp., humidity, barometric pressure), the wind, the site elevation, any slope to the target to the 100th of a degree, and the time between shots. And any other 27 variables you'd care to name. Each rep takes the rifles to a different marksman, and only conducts the test once. What neither any of them or any of the marksman, knows, is that Rifle #1 is an average rifle with factory standard ammo, the second is an accurized rifle with standard ammunition, and the third is an accurized rifle with match-grade ammunition. Group size for each string is recorded, and after all the tests are performed with all the rifles and all the marksmen, the data from each shoot is compiled into a meta-study to see the final results.

That's the kind of meta-analysis you're hoping to perform and analyze.

But if, instead, you hand out 100 rounds to the ten reps, and tell them to go to ten shooting ranges, and record no data other than group size, and just have the first ten people they meet try a 10-round string, and record those results, it's a wee bit different. Because alongside a couple of legitimate riflemen, some of the shooters have never fired any rifle before, some are nearly legally blind, some are likely drunk, and you suspect at least one of the reps threw the ammo into his own garage, and just pencil-whipped phony "data", and turned it in to HQ. The results of those 100 strings are likely not going to be representative of anything useful whatsoever.

Let's see what we have here.

3) Let's Meet the Contestants

The Assclowns In Question (hereinafter AIQ) started with 18,950 COVID studies. They ended up only looking at 34.

We'll get to the details in a moment, but for openers: they threw out 18,916 studies, IOW 99.82058047493404% of them, and based their entire survey conclusions on the results of the remaining 0.17941952506596% they actually looked at.

To put this in perspective, that would be like deciding the 2020 presidential elections on the votes of a population the size of Wyoming, amidst 158M voters. It would be basing the winner of the World Series based on 10 randomly-selected at-bats per team in any season.

Which begs a couple of questions:

If there were 18,950 assorted COVID studies to look at, why did only 34 actually evaluate lockdown effectiveness in curbing mortality?

If any one of those studies were pure poorly-designed and invalid crapola, it would represent only 0.0052770448548812% of all COVID surveys.

But if the same were true for one of the 34, it would be weighted 584 times more heavily.

So how did they chuck all those other studies? They threw out 17,542 based solely on looking at the title, which is generally about ten words or less. So if the title didn't sound applicable, without any further checking, they chucked it. Quelle scientifique!

They threw out another 934 because they didn't measure the effect of lockdowns on mortality, ignoring, yet again, the possibility that they did so because one has no effect on the other, the entire point at issue. Pure scientific genius in action, right here.

That left them 117 studies that they actually, y'know, read. And promptly dumped 83 of them, i.e. all but 34 studies. Because reading more than that would be...work?

4) So, about those "lockdowns" they looked at... 

"Compulsory non-pharmaceutical interventions (NPIs), commonly known as “lockdowns” – policies that restrict internal movement, close schools and businesses, and ban international travel – have been mandated in one form or another in almost every country. We use “NPI” to describe any government mandate which directly restrict people's possibilities. Our definition...include[s] mandated interventions such as closing schools or businesses, mandated face masks etc. We define lockdown as any policy consisting of at least one NPI as described above."

 SRSLY???

Allow me to illustrate:

Sorry for all the short bus kids who could tell us what the windows taste like, but every one of those examples are what the authors of this "study" consider lockdowns.

So, for the Common Core glass-tasters in the crowd:

Who would like to hazard a guess why this sort of "lockdown" might not accomplish...anything?

This is just the level of total horseshit found in the first seven pages, and only 53 more to go.

So keep those "Yeahbuts" coming in the Scratching Post, er, Comments.