Live data from Hacker News

IQ is largely a pseudoscientific swindle

medium.com

351–360 of 400 posts

Re: IQ is largely a pseudoscientific swindle

#351

Earlier quoted context omitted.

You should ask yourself what it would it take to falsify your working hypothesis and then see if there are research results which falsify it.

Longitudinal studies in this area (developmental effects on intelligence) are vanishingly rare. Those that do exist are either extremely limited, ethically suspect, or useless (confounded). The difficulty in answering "why is person/are people the way they are?" in a rigorous, falsifiable way has brought about the existence of entire disciplines dedicated to qualitative analysis as a means of getting closer to answer…

There are natural longitudinal experiments in the form of twin studies. They show pretty conclusively that the largest systematic influence of intelligence is DNA at about 50%. Almost everything else about intelligence is due to random variation in the environment (and not due to all of the things you would think make a difference, like parenting, schooling, class size, economic status, etc.).

The person I was responding to could also just go talk to a bunch of teachers.

Re: IQ is largely a pseudoscientific swindle

#352

Earlier quoted context omitted.

The leftist position isn't that everyone is inherently equal but that everyone should be valued equally and afforded equal opportunity regardless of their race/gender/ability. The idea of an "IQ" measure is too close to the idea of measuring inherent value, which shouldn't be contingent on things like how much nutrition you had as a child. Furthermore, the idea of "IQ" could be considered dangerous because it often s…

The left uses any inequality of outcome, like the unequal representation of minorities and women in the software development industry, as evidence of discrimination without allowing for the possibility these differences are due to different abilities, choices, or desires. And if anyone challenges the belief that all outcomes should be equal, they retreat behind the very different argument you've just made, which near…

Hold on, I'm a lefty - and don't particularly like that brush you're tarring me with. Now before we go on, I do agree with some of your points. The results we observe are driven by many factors and individual choice is one of them. I object to 'positive discrimination' and the like as it's personally unfair. However, I do have that deep troubling feeling that there is bias all over the place. I'll just go with gender, as you mentioned it (and yes, this is all anecdotal). I'm in the UK, working for an Israeli company. The UK office (and US) is a veritable 'sausage-fest' - however Israel and India have a far higher proportion of women doing these technical roles. My point is that these "personal desires" seem to vary by geography, rather than gender. Now, why would this matter, if everybody gets to do what they want? Well, income. Many western industries seem to have massive gender bias within them - and women always seem to end up in the ones that pay less (which for want of a better metric is our best indicator of 'value' we have). Now within a job role, once you balance in hours worked/flexibility/anything-you-care to include the gender pay gap isn't too bad. It's there, but not enough to annoy me. My quibble is that if you/I had two children, one male, one female, equal in every way - the male child is going to earn more.

Re: IQ is largely a pseudoscientific swindle

#353
post #352

Earlier quoted context omitted.

The left uses any inequality of outcome, like the unequal representation of minorities and women in the software development industry, as evidence of discrimination without allowing for the possibility these differences are due to different abilities, choices, or desires. And if anyone challenges the belief that all outcomes should be equal, they retreat behind the very different argument you've just made, which near…

Hold on, I'm a lefty - and don't particularly like that brush you're tarring me with. Now before we go on, I do agree with some of your points. The results we observe are driven by many factors and individual choice is one of them. I object to 'positive discrimination' and the like as it's personally unfair. However, I do have that deep troubling feeling that there is bias all over the place. I'll just go with gender…

That's another area that merits research beyond simply assuming that an unequal outcome implies discrimination, and research to-date isn't entirely supportive of that assumption:

> A massive study that looks worldwide at how national wealth and gender equality affect the choices men and women make ... found greater national wealth and gender equality are tied to bigger differences in preferences between men and women.

1: https://www.scientificamerican.com/article/when-times-are-go...

Re: IQ is largely a pseudoscientific swindle

#354

Earlier quoted context omitted.

No, these are some of the tests I work with. Pop culture turned them into a proxy for IQ or Inteligence etc. They're not. The SAT for example is highly targeted towards one specific goal: predicting level of success academically in the first year of college, and only the first year. They make no other primary research claims about the utility of how the test results should be used, and in dealing with them they are v…

The old SAT correlated more with standard iq tests and can be used for admission into MENSA from what I remember.

The old SAT has a statistically significant correlation, but if memory serves (it might not, it's been a while) the correlation wasn't great, it has something like an r-squared value of 0.4 to 0.5, so a majority of the variance in outcomes was not explained by differing performance on the tests.

Re: IQ is largely a pseudoscientific swindle

#355
post #108

Earlier quoted context omitted.

If I understand you, you are arguing that parents should praise their kids for smartness because it helps kids to understand that they are smarter? But there is not point doing that, if they are smarter they will figure it theirselves. At the same time if parents are praising them for smartness it would lead smartness to become a part of self-esteem and therefore their psychological well being will be tied with their…

> you are arguing that parents should praise their kids for smartness because it helps kids to understand that they are smarter? I’m arguing for truthfulness. If your kid is smart, don’t shy from saying so. If they excel in other ways, point those out. Helping us to understand our strengths and weaknesses is a gift parents can give. > if self-esteem of a person based on working hard it is very good Not necessarily. I…

There is subtle difference between praising and telling kid that he/she is smart. Smartness should be unconditional idea for a person. I believe that 3-9 years old is the best period when parent should tell repeatedly to his kid that he/she is smart, pretty, brave and so on. But unconditionally, for child didn't come to an idea that he/she is good only if he/she is smart, pretty and brave. If he/she does, then she would be unable to say something stupid, behave cowardly, or look not so pretty without believing that she is bad. This means also that parent must never tell to his child that he is stupid, coward, ugly and in any other sense have traits that are not socially desireable. Unconditionally.

All books written by psychologists to parents would tell you that. At least all I've read do.

> But many innovations have been sprung from a combination of intelligence and laziness.

If your child would grow to be completely not lazy person, then either you are genius of practical psychology and education, or your child have some inherent traits that make him not lazy and your efforts made little or no difference. The second is more probable. Moreover I do not believe that smart kid can grow non lazy. If he did he is not smart enough to see how he can cut corners.

I know about innovations, and I did some in the past. I wrote programs for automation but you now, I had never seen success because I was a way too lazy to make this programs usable for someone except of me. My programs die when I change job. I was a way too little motivated to do more than just to write the program. And yes, I never believed that I can work hard for more than a few months. I'm "too smart" for a hard work, if I work hard it means for some deep neural net in my head that I'm not smart enough. I got this idea in the school, where I did nothing and known math better than anyone else, because I had read all math textbooks two years ago just for fun. It was supported by my knowledge that my IQ reach 142 and it is very high. I got this idea and I cannot get rid of it even now, when life keeps proving me that the idea is deadly wrong. And yes, I fear to be stupid, this fear sometimes works even if I managed to stop my impulse to make a stupid thing before anyone noticed.

It is just anecdote, but I'm almost got my bachelor degree in psychology, I'm working with psychotherapist around my problems, and I believe that self-esteem should be based on manageable things, on something that is under control of a person. Smartness is not under control of a person, you can behave on impulse without thinking, you can be tricked by circumstances or other people, you can do a lot of stupid things, while trying to be smart. Moreover there are circumstances when to do something now is more important than to do something smart. So smartness is not a good base for a self-esteem.

Re: IQ is largely a pseudoscientific swindle

#356
post #340

Earlier quoted context omitted.

Does anyone use it like that though? Isn't he arguing with a straw man? Who is using IQ to allocate funding to pop singers? IQ scores aren't used that much and only ever for predicting success of jobs where success is easily measurable (which is what Taleb seems to mean by "obediency-based BS jobs").

Yeah they do, I've seen https://www.worlddata.info/iq-by-country.php linked in discussions about if Germany should welcome refuges from Africa for example. And the gist was that while getting Asian high achievers is a net win for the country, letting refuges from Africa is, because of the linked map, a loss for the country.

Given that not even Taleb disputes IQ effectiveness on the left-end of the spectrum, meaning the low end, letting waves of refugees in from sub-saharan Africa (average IQ below 70 [1][2]) would definitely be a net loss for a first-world country.

His first sentence in linked article starts with: '“IQ” is a stale test meant to measure mental capacity but in fact mostly measures extreme unintelligence'. Why would any developed country want hordes of "extremely unintelligent" immigrants?

Which is why there exist countries with extremely harsh immigration policies that are first-order eugenic. Singapore is one of them.

[1] There are many studies that report similar values, https://www.sciencedirect.com/science/article/pii/S016028960...

[2] Anecdotal but informative: https://undark.org/article/in-south-africa-decolonizing-math...

Re: IQ is largely a pseudoscientific swindle

#357
post #130
post #108

Earlier quoted context omitted.

If I understand you, you are arguing that parents should praise their kids for smartness because it helps kids to understand that they are smarter? But there is not point doing that, if they are smarter they will figure it theirselves. At the same time if parents are praising them for smartness it would lead smartness to become a part of self-esteem and therefore their psychological well being will be tied with their…

But what's wrong with basing one's self-esteem on being smart, if someone is smart?

Because intelligence is an individual trait that is totally not under control of a person. It will not work this way for a long even if person is smart. Especially if she is smart: she will see that to believe that "smart means good" is stupid (what means to be smart? to have good grades? to create a successful startup? to speak smart things? to hear others praise your smartness). So person will risk to start obsessing about smart behaviour, overcontrolling impulses, think twice before opening your mouth, a mistake is a fail that damages self-esteem, and so on... It can be good when you manage to find some unexpected creative and smart idea, but it is fun even without self-esteem: insight in any case is rewarded by brains. But creativity is an ability to generate a whole lot of ideas, stupid and smart, what would you do if your brain stopped to generate any idea in a fear to generate a stupid one?

To be smart is to find non obvious idea, but to find a good idea you need to think a lot, and no rewards while you are thinking. Frustration is the common state of good thinker, and frustration itself seems from inside as inability to solve problem, as a lack of intelligence. You are working hard, but the only visible result is inability of your intelligence to solve the problem. You can fight with fatigue for some time, but if a hard work is not rewarded by you brains as a prove that you are a good boy, then this time can be too short to actually solve the problem.

You should belive that you are smart unconditionally. And if sometimes you are start to doubt it, it should not be a big deal for you. Our society force people to believe that smart is good and stupid is bad, so we are subjects of a narrative that "if you are stupid, then you are bad boy. Go to your room!". There is no need for parents to make special efforts to make this narrative more powerful in your head.

Re: IQ is largely a pseudoscientific swindle

#358
post #4

IQ is undoubtedly overemphasized by some factions, especially those with an axe to grind and a dog whistle to blow, if it helps them feel less guilty about their position in the world or otherwise rationalizes injustice. OTOH there's definitely a difference between people, people who aren't mentally defective, people that are quick on the uptake, sharp and alert, who adapt to new mental models. IQ might be an unscien…

The original tweetstorm and this post aim to show something very specific, namely that IQ misses the mark as a measure of intelligence the further away from IQ 70 one ventures. In other words, stupid is stupid, but smart seems to come in many flavors. The title of the article itself points to this fact, as the word "largely" is inserted. Nassim Nicholas Taleb himself can see the value at the lower levels. If the goal…

His own argument is bust in that in that success and failure are themselves asymetric, so it's a moot point.

"Failing" or "not succeeding" is the default and anyone can do it. In fact, the majority of people will never "make it" and most people even who do semi-well will live comfortable but not amazing lives.

"Success" which he seems to judge as perhaps being a somebody or doing something notable follows a power law distribution. It's rare.

Ok, so most people work the 9-5. And what does Taleb have to say about that? He says "ok, well, yes, if you're a slave then IQ seem to predict you'll do better in that particular game". Right. So it is a good predictor of how well you can expect someone to do when they play the game that the vast majority of people play.

So... In other words it's roughly speaking useful.

Don't even get me started on on his term IYI (intellectuals-yet-idiots). Implying that there are people who are intellectuals that suffer no bounds to their rationality nor any partisan or ideological bias. Right. Only an IYI could even come up with such a concept. I see this all the time. Disagree with someone in academia? Oh, well they are a pseudo- intellectual! Or perhaps you might consider the alternative that you just used pseudo-reasoning to reach your psuedo-conclusion.

He draws an analogy between "best measure" proponents and "Value at Risk" proponents.

Ok. Well let's lay out the tools that will help us analyse that analogy. In the book Shortcut John Pollack describes the 5 things that characterize an effective analogy.

1) uses the familiar to explain the unfamiliar

2) highlights similarities while obscuring differences

3) identifies a useful abstraction

4) tells a coherent story

5) resonates emotionally

It certainly resonates emotionally if you accept the premise that it tells a coherent story. It certainly tells a coherent story that relying on a bust metric will catastrophicly blow up in your face if you accept the premise that it identifies a useful abstraction. It identifies a useful abstraction that a metric might not actually be measuring anything real if you accept the similarities and don't look at the differences.

I'm not part of the finance world and don't have a good enough grasp to deconstruct the analogy enough to figure what the differences are and whether it is a good or a bad analogy, but I wanted to lay out the tools to help people do so.

My gut says thats not a good analogy. If only because people say that well it seems to be the best measure we have and he says that's what the VaR guys said. But then he also says, well yeah if you're talking about the domain in which the majority of people spend the the majority of their lives (the workplace) then yes, it does actually seem to have some predictive power there.

He himself is fooled by randomness.

Re: IQ is largely a pseudoscientific swindle

#359
post #321

Earlier quoted context omitted.

I gave examples elsewhere in this thread that illustrate why more time won't help. There is a large difference between solving the same type of problem, only larger, fast, and using abstractions at a higher level. Do you really think that speed of doing long division predicts ability to solve differential equations?

> Do you really think that speed of doing long division predicts ability to solve differential equations? I never said that. I also fail to see how "solving differential equations" has anything to do with intelligence (with the caveat that pathologically low mathematical aptitude might limit the ability to learn it at all - one aspect here is that "unable to do division" might be a predictor for that). But I believe…

I agree that long division measures some dimension of intelligence. But there are so many dimensions that knowing a coordinate in one dimension does not tell you much about coordinates in other dimensions. Taleb's point is that there is correlation, but it's mostly driven by the left side of the distribution.

"Unable to do long division" usually means that this person cannot solve differential equations. "Able to do long division" tells you very little about that person's ability to solve differential equations.

The speed of solving simple problems does not necessarily predict ability to solve really complex problems well. For example, a chimp named Ayumu easily beats humans in short-term memory tests: https://www.theguardian.com/science/2013/sep/29/chimp-intell...

Re: IQ is largely a pseudoscientific swindle

#360
post #113

Earlier quoted context omitted.

I gave an example above. Imagine you ask second graders to write math tests for college. They might extrapolate from their experience like this "We started working with single-digit numbers, like 2+3 or 4*2. Then we learned to work with two-digit numbers, 23+17 or 17-73. High school math should probably be three-digit numbers, college will be four-digit numbers, and math professor must be able to work with eight-digi…

Your example describes a lack of knowledge, not a lack of intelligence.

Yet the same argument can be made for (any specific aspect of) intelligence.
Post reply on HN