150 sats \ 10 replies \ @SimpleStacker 15h \ on: On These Questions, Smarter People Do Worse Politics_And_Law
Pretty interesting.
I've observed that a lot of smart people are really bad at statistics. I realized it's because in our education system, the smartest kids take AP Calculus not AP Statistics. So it's pretty easy to go through college without having a solid grasp of statistics. Even many Engineers have a poor grasp of statistics, not to mention Doctors who often have to evaluate scientific evidence but are astoundingly bad at it.
Anyone who sees the results matrix at 0:48 of the video and has a statistics education will know how to interpret it... but I bet a lot of highly educated people in our society has never seen such a thing... or if they have it was only for a little while.
This really affected the interpretation of Pfizer's COVID vaccine studies. I think the headline result was that people in the treatment group were 4 times less likely to get COVID than the control group. But the numbers were like 1 in 2,000 vs 4 in 2,000. It shows how you can manipulate people's perceptions based on how the results are presented. And they only tested one outcome, whether or not you got COVID, they didn't test side effects.
(I may or may not be slightly misrepresenting the Pfizer study, since it's been a long time I looked at it. But I think the general impression is accurate, of the initial study that bureaucrats cited to prove the vaccine is effective.)
A lot of smart people are terrible investors.
Didn't Isaac Newton lose a bundle in the South Sea bubble?
reply
That sounds familiar.
The investing thing definitely fits this trend. Smart people think they're smart enough to not make the same mistakes that everyone else make.
reply
That's true, but it looks like it's simple enough for them to figure it out, until it's about something that challenges their beliefs.
I recall a similar finding, which they might have been alluding to in the video, about people with more knowledge of confirmation bias actually being more susceptible to it.
reply
Case in point. My daughter's science teacher who I think has a masters degree in biology or something... she was teaching them about p-values and said that "If the p-value is <5%, then that proves that X affects Y."
I wanted to blow my brains out.
reply
Oof. That is someone who has spectacularly failed to grasp the core premise of statistics (and really science in general).
reply
She's no longer their teacher (she was let go).... but the new one doesn't seem much better. In general, it's been hard for schools to find good science teachers. I think people who are actually good at science have better job opportunities.
reply
people who are actually good at science have better job opportunities
Definitely. Two of the good science teachers I had, were people who had experienced professional success before changing careers. They knew their stuff, but realized they had more passion for teaching.
The other good science teacher I had was just a redneck who loved teaching physics.
reply
First public housing and rent control , now p value hacking!
reply
lmao, you've been paying attention :)
Yes, the academic adventures of my kids are interesting for me to behold.... and she actually goes to what's considered a very good school for the area.... explains why so many of American society's "elites" are actually quite foolish. I just hope that my influence can counter some of the bad educational programming she's getting.
reply
The correct answer is "if p value is < 5% then your findings can be published in a scientific journal"
reply