There are people who desperately want this to be untrue🧵
One example of this came up earlier this year, when a "Professor of Public Policy and Governance" accused other people of being ignorant about SAT scores because, he alleged, high schools predicted college grades better.
The thread in question was, ironically, full of irrelevant points that seemed intended to mislead, accompanied by very obvious statistical errors.
For example, one post in it received a Community Note for conditioning on a collider.
But let's ignore the obvious things. I want to focus on this one: the idea that high schools explain more of student achievement than SATs
The evidence for this? The increase in R^2 going from a model without to a model with high school fixed effects
This interpretation is bad.
The R^2 of the overall model did not increase because high schools are more important determinants of student achievement. This result cannot be interpreted to mean that your zip code is more important than your gumption and effort in school.
If we open the report, we see this:
Students from elite high schools and from disadvantaged ones receive similar results when it comes to SATs predicting achievement. If high schools really explained a lot, this wouldn't be the case.
What we're seeing is a case where R^2 was misinterpreted.
The reason the model R^2 blew up was because there's a fixed effect for every high school mentioned in this national-level dataset
That means that all the little differences between high schools are controlled—a lot of variation!—so the model is overfit, explaining the high R^2
This professor should've known better for many reasons.
For example, we know there's more variation between classrooms than between school districts when it comes to student achievement.
College students make or are forced to make suboptimal choices about the times their classes take place🧵
For students who register for 8AM classes, about a third wake up after class starts, and almost 40% wake up too late to get to class on time.
People's internal rhythms aren't things they just choose, they're somewhat out of their control because they're synced up with day-night cycles.
Consider this, showing the amount of time 8AM class-takers sleep on school days vs weekends (gray), measured through logins at school.
If you compare those 8AM class-takers to 9AM students, you see that the ones who registered for 9AM classes sleep longer, but both sleep similar lengths on weekends.
"A reminder: the US has the #1 most expensive healthcare system in the world, yet we rank roughly #42 in life expectancy."
Brian Thompson's murderer wrote that in his manifesto.
Both claims are bad. The first one, because America spends the most on healthcare because it's rich:
I don't mean Americans pay higher prices for the same amounts of care, but that Americans consume much higher volumes of care. They do more check-ups, get more screenings, take more tests, dose more drugs, get more surgeries... and so on!
You can predict spending from volumes:
Regarding the second claim, Americans have shorter lifespans because they're fat, violent, and reckless, not because of things that the health system can control.
And, if anything, when it comes to the things the health system actually controls, they generally do better!
In Medicare Advantage, the government encourages insurers to pursue perverse incentives.
These see them overdiagnosing patients for conditions they often do not have, and which they don't recommend treating, because it means they can send the government a bigger bill.
Insurers have tried to claim that they don't do this, and that this data misrepresents the care they provide to patients they diagnose with particular conditions (like HIV!) because COVID disrupted care, but they're lying and it's too obvious.
Medicare Advantage does beat FFS on performance and on cost, and it certainly leads to more care for tons of people, but like every major program in medicine, it gets defrauded to the tune of tens of billions of dollars.