Hacker News Board Hacker News, browsed like a classic message board
Live data from Hacker News

Why The Pentagon run the best schools and the safest nuclear program

governance.fyi

Post reply on HN
Print view View original

71–80 of 102 posts

Re: Why The Pentagon run the best schools and the safest nuclear program

#71

Post by whimsicalism »

Dylan16807 wrote: post #68
zevon wrote: post #54

Earlier quoted context omitted.

Why do you insist on saying that I "don't believe" in genetic components when I've literally said the opposite? The people who wrote the stuff on the Wikipedia site I was provided with and their (researcher-)sources seem to try to tell you and me both "hey, this is an interesting field of study but it's very complicated, many genes are involved, we are far from understanding them or being able to model them, be very…

> "hey, this is an interesting field of study but it's very complicated, many genes are involved, we are far from understanding them or being able to model them, be very careful with interpreting correlations and for (m)any practical purposes (such as thinking about how to structure educational environments), you really should consider quite a lot of things not directly related to genetics." I'll say the same thing a…

Absolutely agreed. I got bogged down in the genetics portion, but it is not actually a necessary component of the argument I'm trying to make - merely that kids are like parents.

Re: Why The Pentagon run the best schools and the safest nuclear program

#72

Post by whimsicalism »

zevon wrote: post #57
whimsicalism wrote: post #55

Earlier quoted context omitted.

you’re not arguing in good faith and now you’re motte-baileying. you said: > Correlations between socioeconomic status and success of one's offspring in educational systems don't mean that intelligence is inherited in the genetic sense. If you're seriously arguing this, you're very close to flirting with eugenics and the like. the obvious reading is that you do not believe in a genetic component to intelligence - and…

My original answer was a condensed and far from comprehensive one-sentence reply to another condensed one-sentence-reply (that included the phrase "highly heritable" which is how the whole genetics argument started). Why is what you apparently perceive this original one-liner to mean so important to you? I've expanded on the points I was trying to make quite a bit. And again: The researchers who look at those things…

> Why is what you apparently perceive this original one-liner to mean so important to you?

Well because you basically accused most people of being eugenicists simply for believing something that is most likely true and clearly implied a strong position that you are now retreating from. It's clearly an incendiary one-liner where previously the conversation was not so.

> The researchers who look at those things seem to be the ones telling us that the relationship between intelligence and genetics is complicated and many, many non-genetic factors are in play, no

There are massive biases in academia that encourage researchers to hedge results like this. When you ask anonymously, the answers & beliefs are clear.

> Snyderman & Rothman (1987/1988) — mailed survey to ~1,020 academics; 661 replies. Experts overwhelmingly agreed that IQ has substantial within-group heritability, and among those willing to give a number, the average estimate was ~60% for U.S. populations. Also https://pmc.ncbi.nlm.nih.gov/articles/PMC4804158/ which is going to be a lower bound because it focuses on international differences.

Adoption studies pretty clearly upper bound the amount that these complicated non-genetic/non-prenatal factors can be causing differences in tested adult intelligence among Americans.

> I'd politely ask you to refrain from insulting my good faith.

Again, you started your entry into this conversation by leveling accusations of eugenics. The responses you get are going to be tinged by that.

> how you come up with that clear of a statement about smart parents and their non-externally-influenced child, how one would approach that as a research question/design

Adoption studies can provide an upper bound (excluding pre-natal environment). Also GWASs paired with mendelian randomization can provide a lower bound.

Re: Why The Pentagon run the best schools and the safest nuclear program

#73

Post by thaumasiotes »

nitwit005 wrote: post #56
thaumasiotes wrote: post #38

Earlier quoted context omitted.

You qualify based on the AFQT, which is a composite of four ASVAB subtests. Specifically, the Word Knowledge, Paragraph Comprehension, Arithmetic Reasoning, and Mathematics Knowledge subtests. Sample questions: (identify the best synonym for the underlined word) His record provides no reason for _apprehension_. [ ] anxiety [ ] change [ ] enjoyment [ ] endorsement According to the passage, some artists work with titan…

You'll notice, that is basically your ability to read and do math, which is rather different than the content of a typical IQ test. People have apparently done research on adding more IQ test sections in: https://www.sciencedirect.com/science/article/abs/pii/S22113...

> that is basically your ability to read and do math, which is rather different than the content of a typical IQ test.

No, it isn't different. Do you know what an IQ test looks like?

Re: Why The Pentagon run the best schools and the safest nuclear program

#74

Post by tptacek »

Dylan16807 wrote: post #31
zevon wrote: post #28

Earlier quoted context omitted.

Correlations between socioeconomic status and success of one's offspring in educational systems don't mean that intelligence is inherited in the genetic sense. If you're seriously arguing this, you're very close to flirting with eugenics and the like.

> If you're seriously arguing this, you're very close to flirting with eugenics and the like. Please don't be so eager to reject eugenics that you end up being anti-science. The idea that some percent of intelligence is genetic is entirely reasonable, not something to refuse to consider. And there's good evidence too: https://en.wikipedia.org/wiki/Heritability_of_IQ

What do you think "the heritability of IQ" means? It seems from the thread that you believe it's genetic causation of intelligence. Is that what you're claiming?

Re: Why The Pentagon run the best schools and the safest nuclear program

#75

Post by Dylan16807 »

tptacek wrote: post #74
Dylan16807 wrote: post #31

Earlier quoted context omitted.

> If you're seriously arguing this, you're very close to flirting with eugenics and the like. Please don't be so eager to reject eugenics that you end up being anti-science. The idea that some percent of intelligence is genetic is entirely reasonable, not something to refuse to consider. And there's good evidence too: https://en.wikipedia.org/wiki/Heritability_of_IQ

What do you think "the heritability of IQ" means? It seems from the thread that you believe it's genetic causation of intelligence. Is that what you're claiming?

I think that's how zevon interpreted it, so that's what I responded to.

Personally I would include other methods. And for the argument about schools method doesn't really matter.

Edit: And I don't know if genetics are the biggest factor in single generation inheritance, especially at a younger age, but I do think they're a reasonably significant factor after looking at various estimates.

Re: Why The Pentagon run the best schools and the safest nuclear program

#76

Post by tptacek »

Dylan16807 wrote: post #75
tptacek wrote: post #74

Earlier quoted context omitted.

What do you think "the heritability of IQ" means? It seems from the thread that you believe it's genetic causation of intelligence. Is that what you're claiming?

I think that's how zevon interpreted it, so that's what I responded to. Personally I would include other methods. And for the argument about schools method doesn't really matter. Edit: And I don't know if genetics are the biggest factor in single generation inheritance, especially at a younger age, but I do think they're a reasonably significant factor after looking at various estimates.

Right, just so we're clear that heritability isn't genetic determinism; it's a correlation statistic. All sorts of things are heritable that are absolutely not fixed by genetics. And there are things fixed by genes that aren't meaningfully heritable!

Re: Why The Pentagon run the best schools and the safest nuclear program

#77

Post by zevon »

Dylan16807 wrote: post #68
zevon wrote: post #54

Earlier quoted context omitted.

Why do you insist on saying that I "don't believe" in genetic components when I've literally said the opposite? The people who wrote the stuff on the Wikipedia site I was provided with and their (researcher-)sources seem to try to tell you and me both "hey, this is an interesting field of study but it's very complicated, many genes are involved, we are far from understanding them or being able to model them, be very…

> "hey, this is an interesting field of study but it's very complicated, many genes are involved, we are far from understanding them or being able to model them, be very careful with interpreting correlations and for (m)any practical purposes (such as thinking about how to structure educational environments), you really should consider quite a lot of things not directly related to genetics." I'll say the same thing a…

Again, the person I was originally replying to called intelligence "highly heritable". That does mean a genetic argument and I replied to that and not a generic assertion that there are mechanisms in play that have influence on the expression over generations.

Re: Why The Pentagon run the best schools and the safest nuclear program

#78

Post by Dylan16807 »

tptacek wrote: post #76
Dylan16807 wrote: post #75

Earlier quoted context omitted.

I think that's how zevon interpreted it, so that's what I responded to. Personally I would include other methods. And for the argument about schools method doesn't really matter. Edit: And I don't know if genetics are the biggest factor in single generation inheritance, especially at a younger age, but I do think they're a reasonably significant factor after looking at various estimates.

Right, just so we're clear that heritability isn't genetic determinism; it's a correlation statistic. All sorts of things are heritable that are absolutely not fixed by genetics. And there are things fixed by genes that aren't meaningfully heritable!

Of course it's not determinism but it is statistically causal. The point is the distribution of the student body is going to be different in a pretty significant way.

Re: Why The Pentagon run the best schools and the safest nuclear program

#79

Post by zevon »

whimsicalism wrote: post #72
zevon wrote: post #57

Earlier quoted context omitted.

My original answer was a condensed and far from comprehensive one-sentence reply to another condensed one-sentence-reply (that included the phrase "highly heritable" which is how the whole genetics argument started). Why is what you apparently perceive this original one-liner to mean so important to you? I've expanded on the points I was trying to make quite a bit. And again: The researchers who look at those things…

> Why is what you apparently perceive this original one-liner to mean so important to you? Well because you basically accused most people of being eugenicists simply for believing something that is most likely true and clearly implied a strong position that you are now retreating from. It's clearly an incendiary one-liner where previously the conversation was not so. > The researchers who look at those things seem to…

No, I said one is "close to flirting with Eugenics" wich is rather not the same than accusing anybody of being an Eugenicist and I stand by that point. However, you and the other person insisting on (mis-)interpreting my original one-liner now seem to do the "retreating" and to say that the post I was replying to somehow "clearly" was about the generalized notion of kids being like their parents instead of being very specifically about genetics.

The study you linked is interesting but its results are far from "clear" (see its discussion section but that's probably also just bias and hedging or whatever) and it does have fuck all to do with your proposed thought experiment of a Kaspar-Hauser-like child. Even less so with your confident prediction of how a Kasper-Hauser-like child would turn out. I think you probably know this yourself but these kinds of predictions are something scientists would very, very rarely do - because they know the limitations of their work.

I'm kind of weirded out by this exchange, people here rather confidently express quite a bit of stuff that goes against years and decades of training I received when I became a scientist and I think I'll stop replying now. That was the recommendation of a colleague - who actually is a geneticist - I showed this thread to over coffee as well.

Re: Why The Pentagon run the best schools and the safest nuclear program

#80

Post by Dylan16807 »

zevon wrote: post #79
whimsicalism wrote: post #72

Earlier quoted context omitted.

> Why is what you apparently perceive this original one-liner to mean so important to you? Well because you basically accused most people of being eugenicists simply for believing something that is most likely true and clearly implied a strong position that you are now retreating from. It's clearly an incendiary one-liner where previously the conversation was not so. > The researchers who look at those things seem to…

No, I said one is "close to flirting with Eugenics" wich is rather not the same than accusing anybody of being an Eugenicist and I stand by that point. However, you and the other person insisting on (mis-)interpreting my original one-liner now seem to do the "retreating" and to say that the post I was replying to somehow "clearly" was about the generalized notion of kids being like their parents instead of being very…

> your proposed thought experiment of a Kaspar-Hauser-like child

Oh my god this is the most malicious possible reading of "remove all environmental factors".

They're talking about making the environments match, for fuck's sake.

The bulk of your comments are arguing that heritability is very complex, which is completely compatible with the words "highly heritable". And you still haven't explained why the term "eugenics" was relevant to anything anyone else said. If it's something about race, a superficial similarity across millions of people in shifting groups has very little to do with the correlations between child and parent that share 50% of their genes, but even if those were the same the comment still didn't say anything that got anywhere near eugenics!!

> the post I was replying to somehow "clearly" was about the generalized notion of kids being like their parents instead of being very specifically about genetics

I didn't use the word "clearly". You're misquoting now too?

And I still believe they meant the entire complex mess you're talking about, yeah. I think you have zero justification to barge in and say it's a complex issue, and the person making a single sentence comment must have meant the most simple possible version, there's no way they were referring to the entire complex issue already without your help, in a context where the distinction doesn't even matter.

Even if you're done replying I hope you see this: If you were actually talking in good faith you're doing a very bad job at giving anyone the benefit of the doubt for how they word things.

Post reply on HN