A funny little thing happened the other day. I was golfing with my brothers and I hooked my tee shot towards some back yards. When I couldn’t find my ball my brothers came over and said “there you are” pointing to someone’s back yard. My ball had hopped through their fence. I noticed every single post had a Trump 2024 flag and said “are those Trump flags? Fuck them.” I then noticed the old couple sitting on some chairs on their back patio. I rode back out to the fairway and took a drop there.
4 Likes
CLT applies if you sum independent random variables.
No idea how IQ tests work but assuming the score is the number of correct answers, distribution would be roughly normal if there were no correlation between getting question x right vs getting question y right.
So for instance if an IQ test consists of say 20 very different categories which measure completely different skills, then distribution will be close to Gaussian, but I doubt that’s the case ?
If there’s strong correlation to getting positive answers on pairs of questions then I think the distribution will be more spread out than a Gaussian
You’re out there on a crisp fall morning with RFK doing the falconry and he’s telling you all about the particular bird and why he loves him and how he’s such a special and smart falcon and then BOOM FEATHERS EVERYWHERE Don Jr fucking smoked it with his 12 gauge shotgun and then him and Eric are high fiving and posing with it’s exploded carcass.
5 Likes
Boy you really cannot post funny clips from Kentucky Fried Movie without watching them first. Thanks, Biden.
But I mean, ignoring the test, “intelligence” itself (whatever that means) is probably due to a sum of all kind of independent random variables the same way height and blood pressure is, right?
1 Like
9 Likes
Indeed, “intelligence” is not an actual trait, like “athleticism” is not an actual trait, it’s a complex melange of an undefined number of underlying related and unrelated variables. I assume that any such loosely defined “trait”, when operationally defined into some score, will likely be normally distributed.
Define athleticism as “vertical leap” and it will be largely normal (though with perhaps some clustering, like 0" because of non ambulatory people). I presume most human “traits” are like this, from BP, lean muscle mass, height, empathy, rudeness, longevity, daily energy expenditure, average bowel movement size, etc.
https://x.com/Jose_Pagliery/status/1844454542857097611
https://x.com/Jose_Pagliery/status/1844455072895471980
Having trouble finding the right thread. UP yours, discourse.
5 Likes
We got him.
13 Likes
IQ is the made up distribution. It’s normal because they’ve specifically designed it to be normal. That doesn’t imply that intelligence, to the degree it is quantifiable, has to be normal.
It’s like if I took football scores and translated them into something I made up like “FQ” that I specifically designed to be normal. You’re asking why I expect it to be normal.
I have people close to me who ate measurably leaps and bounds above me in terms of intelligence still fully aboard the Trump train, so it’s clearly something other than raw intelligence at work here.
It’s the size of their amygdalas. Raw intelligence is no match for fear.
That or they’re rich and want a tax break and don’t give a shit about other human beings.
3 Likes
I’d have put the etc one term earlier.
3 Likes
“Smartest” friend of mine from college is anti-trans and therefore I assume voting Trump.
Yeah, I think the anti-trans prejudice is the main reason Harris/Walz are gonna lose.
Replace bolded with “combination” and it’s probably close to correct, but “sum” has a precise meaning, i.e. you’re assuming an underlying linear structure according to which all these variables interact, and I don’t really see any reason why that should be the case ?
Granted that there’s a lot of CLT’s in probability theory, but there’s also a lot of other distributions that arise universally in more complicated contexts (e.g. semi-circle law for spectrum of large matrices,…)
Also a more obvious point is that a function of Gaussian is not going to be Gaussian in general (unless it’s linear, or so specific that it again falls into a CLT), so if we’re not “ignoring the test”, then even if the underlying “intelligence” is Gaussian, there’s no reason for the test result itself to be Gaussian.
(I think this is my first time participating in an IQ derail…I guess it had to happen eventually lol…sorry everyone)
Nah I don’t think so, anti trans stuff has continually been a loser electorally
1 Like