arthwollipot
Observer of Phenomena, Pronouns: he/him
And there it is. I knew we'd be breaking out The Bell Curve at some point.
Posted By: jimbob
Last edited by a moderator:
-
Due to ongoing issues caused by Search, it has been temporarily disabled
-
You may need to edit your signatures. When we moved to Xenfora some of the signature options didn't come over. In the old software signatures were limited by a character limit, on Xenfora there are more options and there is a character number and number of lines limit. I've set maximum number of lines to 4 and unlimited characters.
Split Thread The validity or otherwise of a generalised measure of intelligence
- Thread starter arthwollipot
- Start date
This seems to indicate formal education does have a beneficial impact on intelligence:
"The results reported here indicate strong, consistent evidence for effects of education on intelligence. Although the effects—on the order of a few IQ points for a year of education—might be considered small, at the societal level they are potentially of great consequence. A crucial next step will be to uncover the mechanisms of these educational effects on intelligence in order to inform educational policy and practice"
https://www.ncbi.nlm.nih.gov/pmc/articles/PMC6088505/
Trausti
Graduate Poster
- Joined
- Apr 4, 2023
- Messages
- 1,799
That doesn't really address college attendance. I don't disagree that education can have a long-term benefit (I make sure my kids do their homework, and then some). But the mere occassion of attending university doesn't seem to move the needle much. The average IQ of university graduates has been in steady decline.
Well college is education, if Mr Average goes for four years then going by the study linked above then Mr Average's IQ could raise by up to 8 points. Sounds a pretty significant gain when you look at the disparity between skin colours.
Trausti
Graduate Poster
- Joined
- Apr 4, 2023
- Messages
- 1,799
But that study doesn't say that. The testing was done mostly on children. And it does not say 8 IQ points. Again, I don't doubt education interventions can have a postive affect on achievement. I'm simply stating that college has little effect on intelligence. Many founders of today's top companies were college dropouts. College is mostly just about credentialing.
Last edited:
It states IQ rose by a few points for each year of education.
arthwollipot
Observer of Phenomena, Pronouns: he/him
I've yet to see convincing evidence that IQ measures anything more significant than a person's ability to do IQ tests.
Last edited:
Weird that it correlates so highly with other life outcomes like job performance and income, then. It seems weird that taking a test when you're 10 that measures "ability to take this test" could predict your income as an adult, as well as many other important life outcomes (years of education is another, for instance, and yes, "income controlling for years of education" is still strongly correlated, so they're not just the same measure).
arthwollipot
Observer of Phenomena, Pronouns: he/him
I am skeptical of such claimed correlations.
Norman Alexander
Penultimate Amazing
You seem to be making the mistake of correlating high IQ with personal and/or commercial success in an English-speaking western society like the USA, UK or Europe.
High IQ westerners taking an IQ test in, say, Turkiye will do very poorly if their knowledge of the the country, the language, and Turkish practices and culture is poor or non-existent.
High IQ westerners will usually perish quickly in environments where so-called low IQ natives can easily thrive. They would consider such westerners to be extremely dumb, low IQ.
So you do understand that IQ tests are usually geared to the society and environment of the people taking them? As Arth said, an IQ rating is really just proof of the ability to complete an IQ test.
Cosmic Yak
Philosopher
Does high IQ not correlate to success in non-western societies?
Struggling with a test written in a language you don't speak, or one that is not your first language, is not a sign of low IQ.
There has been a fair bit of controversy about whether or not IQ tests have a cultural bias. It seems that the current consensus is that socio-economic factors are more significant that any pro-white, pro-western bias.
https://sintelly.com/articles/are-intelligence-tests-biased/
https://www.sciencedirect.com/science/article/abs/pii/0160289679900138
https://nrcgt.uconn.edu/wp-content/uploads/sites/953/2015/04/rm04204.pdf
You seem to be mistakenly equating local knowledge with IQ. That I do not know how to trap a tapir does not make me stupid- just ignorant of the skills needed for survival in the Amazon jungle. Ignorance can be remedied by education, but low IQ is (I believe) rather more fixed.
As I have shown, that's really not true. It used to be thought that there was a cultural bias in IQ tests, or even that the very act of making someone take a test was an act of cultural bias. This is not the case.
I thought the best predictor of your future earnings was your parents' earnings?
Cosmic Yak
Philosopher
That was before Liz Truss.
Not surprisingly your parents' IQ is also a pretty good predictor of your IQ.
But anyway, my claim isn't that IQ is the only thing that influences life outcomes. I don't even claim that it captures everything that we mean by "intelligence". I only claim that it's not just a measure of "skill at taking IQ tests". It's actually measuring something meaningful. Again, this doesn't mean it measures everything that's meaningful.
Are you skeptical that the correlations exist, or of the interpretation of what those correlations mean?
The correlations themselves are very well established. But a more sophisticated critique might ask why those correlations exist. Are people who are good at taking IQ tests simply better at concentrating on a particular present task, for instance? If so, perhaps that skill is the thing that IQ tests are measuring, rather than actually measuring intelligence, in which case it would be the ability to focus/concentrate that is also applicable to job performance and income.
(Re: this particular question, as I understand it it's been tested by researchers and shown to not be the actual root of performance on IQ tests, but maybe you have something like this in mind that hasn't been ruled out)
IQ is a bedeviled subject, and not likely to become less so anytime soon; or ever, in my weary layman's opinion.
My mother taught country -- hell, frontier-- school in Wyoming back in the 1920s and 30s, when IQ testing was fresh and green. Her schoolhouse was once visited by a brisk young lady graduate student who IQ tested all 22 students (1st through 8th grade) and then explained the results to my mother, confidently telling her which children were worth encouraging and which ones were, well, not suitable. My mother, who was a math major in college, listened with interest and thanked the young lady. Telling the story decades later, she still laughed: the test scores had a remarkable ZERO CORRELATION with the individual kids' abilities and general brightness
in my mother's assessment, after teaching them individually over a long winter, one on one tutoring in every case.
So it was an early, pretty basic IQ test, administered just once, vs one country school teacher's subjective evaluation. Which do you rely on?
Zero correlations don't occur very often in nature, I think. But a 1920s approach to a fluid -- and partly subjective -- problem like IQ was early days.
It's still early damn days, if you ask me.
My mother taught country -- hell, frontier-- school in Wyoming back in the 1920s and 30s, when IQ testing was fresh and green. Her schoolhouse was once visited by a brisk young lady graduate student who IQ tested all 22 students (1st through 8th grade) and then explained the results to my mother, confidently telling her which children were worth encouraging and which ones were, well, not suitable. My mother, who was a math major in college, listened with interest and thanked the young lady. Telling the story decades later, she still laughed: the test scores had a remarkable ZERO CORRELATION with the individual kids' abilities and general brightness
in my mother's assessment, after teaching them individually over a long winter, one on one tutoring in every case.
So it was an early, pretty basic IQ test, administered just once, vs one country school teacher's subjective evaluation. Which do you rely on?
Zero correlations don't occur very often in nature, I think. But a 1920s approach to a fluid -- and partly subjective -- problem like IQ was early days.
It's still early damn days, if you ask me.
Thermal
August Member
Have you ever taken one, or studied the administration of one? I've taken two as an adult, the SB and the WAIS. Pretty comprehensive barrage of a spectrum of intellectual measurements, from how quickly you can assemble specific shapes from blocks, to your grasp of analogies. People who perform well normally show a high degree of success with other intellectual activities, and vice versa.
I'd concede that many of the verbal measurements have a lot to do with exposure to a wide range of vocabulary, which a person with poor schooling would not do as well on, just because they weren't familiar with the terms. My wife, administering IQ tests to young children, finds that they might not recognize the word "television" right away, but recognize "the TV" perfectly. Yet the testing requires that you use "television", so some of these kids score lower than they should.
arthwollipot
Observer of Phenomena, Pronouns: he/him
See, the problem is in assigning all of these different kinds of reasoning and intellectual capability a single number. It recognises neither strengths nor weaknesses. It just smears out everything and represents a person's entire intellectual capacity with a single parameter.
It measures nothing precisely other than a person's ability to do IQ tests.
"all these different kinds of reasoning and intellectual capability" are highly correlated with each other. Why? There seems to be something underlying the correlation between those different aspects of cognitive ability. Is it intelligence? That word seems to have a lot of baggage. Researchers call it g, and try to design tests to best measure that underlying thing. They can't perfectly measure it, but the reason for measuring "different kinds of reasoning and intellectual capability" is exactly because doing so gives a better measure of the underlying thing that leads all those abilities to be correlated than only measuring one. IQ tests aren't perfect at this, IIRC they are correlated with g at about 0.8, but that's still pretty good.
But as I said, what IQ tests are measuring is certainly not limited to "ability to do IQ tests". If it were, they wouldn't be correlated with so many other things.
They certainly don't measure the only things that matter or anything like that. That doesn't mean that they measure nothing.
Thermal
August Member
Well...yeah. What would you expect it to be? It's an approximated number showing where the person's intellectual performance range is. For a more detailed battery, a more detailed breakdown is used, showing strengths and weaknesses in different areas.
What exactly do you think an IQ is supposed to represent, and how does it fall short in your estimation?
...right, and an IQ test is designed to approximate intellectual strength. So...seems to be a good indicator.
arthwollipot
Observer of Phenomena, Pronouns: he/him
I've already said how it falls short. It doesn't take into account strengths and weaknesses - it smears them all out into a single parameter. I'm great with vocabulary and language, but hopeless with mental arithmetic and spatial perception. Why should my capacity with one be brought low by my lack of capacity with the other? These two parts of the IQ test are, for me, not correlated with each other at all.
And I'm not alone in my skepticism. Stephen Jay Gould famously compared IQ testing to craniometry in validity and usefulness. Stephen Hawking, when asked his IQ, said that "people who boast about their IQ are losers". My own IQ is undoubtedly lower than both of theirs.
But even if we do grant the "so many things" that high IQ is reportedly correlated with, so what? What do we do about that? How do we use that information? How does it help us?
This suggests that IQ isn't a perfect measure, not that it's not a useful one. Also, IQ is attempting to measuring something underlying both "wordcels" and "shape rotators" (to use the meme). If there are other factors that impact those abilities (and their clearly are), then by using a test that scores many different things that are impacted by that underlying thing we can start to measure that underlying thing and minimize the impact of those other confounding factors.
Here's an analogy: Imagine we want to measure the quality of people's shoes but for some reason we can access them directly. So we have everyone run a race. Those with better shoes will be more likely to run faster. But some people are just better runners. So we also have them walk a balance beam, those who have better shoes are likely to be able to walk further before they fall off. But some people have better balance. So we also hike a gravel road and examine their feet after. Those with better shoes will end up with fewer blisters. But some of them might have had more callouses. So we also...
The point isn't that these tests will all be highly correlated in each individual. The point is that they will all be influenced by the quality of the shoes, and by doing different sorts of tests we can eliminate some of the other things that any particular test would also be testing for.
When you do a vocabulary test, you might be testing for g, but you're also getting an impact based on how many books that person reads. So we also do a shape rotator test. That's not influenced by how many books a person read. We also do a test where you are given a long number and then have to repeat it back in reverse order, and that's not influenced by shape rotation ability, but it is also influenced by g. And by doing all these different sorts of tests, we can start to get at the common thing that influenced all of them.
Hawking is a physicist and Gould a biologist, but the study of IQ is an area of psychology. I don't think that the skepticism of either of them is particularly enlightening of a subject that isn't their field. That doesn't mean that the arguments that Gould brings forth shouldn't be addressed on their merits, but there's no valid argument from authority here.
That's an interesting and valid question, but it's also unrelated to the question of fact. Do you pose it because you want to discuss the usefulness of the measure of IQ, or to suggest that we shouldn't study it?
At the moment I would say it is like Roboramma's analogy we do seem to be measuring something, but exactly what is rather ill defined. You can't say it is intelligence because that becomes a circular definition, i.e. what does an intelligence Q test measure, it measures intelligence, what is intelligence, it is what an intelligence Q tests.
And of course this is again another diversion for the racists, a tactic they use time and time again to try and pretend that science supports their belief in distinct human races which of course always boils down to being able to claim that their "the black race" is inferior/more violent/less civilised and so on.
In general, it can help by reducing unnecessary effort both for the individual and society.
I realise that there are those who consider highlighting differences in ability between individuals to be "unfair" - and insofar as there are such differences then it is indeed "unfair" that those differences exist, that there are those who are less able to navigate the complexity of modern life than others. But those differences do exist, however unfair, and it seems sensible to allocate resources and activities to those who can make best use of them. How much more unfair is it to ask someone to (eg) undertake a course of study in a subject in which they will almost certainly fail than to point that out and lead them into a subject in which they can succeed.
Even the phrase "from each according to their ability" suggests ensuring each has a task in which they can give their best.
That said, as you point out, a single IQ figure is still an excessively crude measure of ability when there are those who excel in some areas but not in others. All that suggests is more complex assessments should be used.
And no I haven't forgotten that IQ tests, even more complex ones, don't give a measure of a person as a human being, but then I never thought they would.
HOWEVER
Returning to the thread subject, it seems to me the whole IQ thing is a red herring. There is no test that measures truly innate cognitive ability and I don't expect there to be one in the near or far future. Quite apart from whether it would show any differences at all, those differences would have to be proven to have effects, and those effects would have to be present wherever those with the differences were present. Not gonna happen...
Yes there have been some genetic changes as the human race has spread across the planet, but AFAIK they have all been superficial physical changes with the exception of some differing succeptibility to disease. None have been detected that I know of at the depths of inate behaviour, such as an inate propensity for violence - and even should one be found in the future, it would also need to be shown to produce physical effects wherever such genetic variation was found to be of any use whatsoever.
And so we return to sociological pressures, particularly in certain situations, so, what are they, and just as importantly, what is the mechanism by which those pressures lead to violence. Finding such a mechanism would be the key to finding a way to break that link.
Thermal
August Member
Right, but what an IQ test is measuring is not "IQ test stuff". It's measuring actual cognitive aptitude for problem solving and understanding of abstractions and... pretty much everything that would reflect what intelligence is. A high IQ score is not hugely meaningful as a number, except to show that "this person performs well at virtually anything you throw at them".
Hence my link to the Cambridge research upthread. A poor quality school will not deliver a wide vocabulary, which translates to scoring poorly on verbal tests. The aptitude may be there, but won't express accurately on a standardized test. A good way to look at it is that the high and low scores are very accurate and solid predictors, and the middle a bit of a mish-mash of people who do well at some things and not at others, then get mad because they think their areas of excellence are being underrated and they shouldn't be "average". But that's kind of what an intelligence quotient number is looking for- all around intellectual performance.
Eta: the tester will note that a score should be higher based on one factor or another, and can even invalidate a score based on an obvious limited vocabulary but excellence in areas where there is less reliance on certain extrrnal information. Testers are fully aware of this, and are actually trained to recognize it. They are not as dumb as laypeople think/assume they are.
Last edited:
Thermal
August Member
An IQ score is not supposed to be some kind of Greatest Hits reel catered to you. It's supposed to reflect how well you perform in a variety of intellectual tasks. You do great in some, and not so hot in others? That'll put you in the middle of the pack. This doesn't make straightforward sense? The specialized areas are not intended to be held up to say "here's where arthwollipot really shines at his best".
I agree with Hawking, as would any test administrator. There's no reason to boast about a high score anymore than boasting about your height or eye color. But that doesn't mean Hawking dismisses the test score itself, as you do. He is blasting blowhards that over value the number.
It's a great barometer in schools to determine whether a child needs special services, and the test and score is indeed part of the barrage in creating an IEP. It can identify high performers who may perform poorly in mainstream classes because they are bored/unchallenged, and curricula can be tailored to help them excel. Just for a start, before adulthood is even reached.
arthwollipot
Observer of Phenomena, Pronouns: he/him
The flip side of this is that it can be used to profile and to discriminate.
Thermal
August Member
Well, that's true, if someone is malicious. Many swords are in fact double edged. Pretty sure my wife is a fan though, as it very often has indicated a student was not cognitively impaired, but learning disabled through their environment.
For adults, I'd agree with you that a score can be misleading, or not valuable at all. Being at the very high or low ends are a helpful indicator of ability, but as you say, are not screaming good Indicators of how sharp a given person is in their preferred wheelhouse.
An electrical engineer buddy of mine had a problem with teaching programs in-house at the local tech center. Seems the guys who wrote the programs couldn't explain them to anyone else, so they couldn't get anyone to run them. So they enlisted a couple high school teachers to learn them, then the teachers taught the operators and it all worked out. Funny thing was, the teachers had no clear idea what the programs did, but with language being their strong suit, they were able to convey the workings of the program without really understanding what it was doing in practice, while the operators did. Everyone has their strong suits.
Trausti
Graduate Poster
- Joined
- Apr 4, 2023
- Messages
- 1,799
Gould was an unapologetic Marxist and admitted his notions were guided by this ideology. Not suprising he'd pooh-pooh anything that suggested human individualism.
Last edited:
Trausti
Graduate Poster
- Joined
- Apr 4, 2023
- Messages
- 1,799
Whatever it is measuring, that measure is one of the most heritable traits. It's as heritable as height. We are not suprised when children look like their parents; so why would we be suprised that they are as a smart and act like their parents?
Last edited:
Trausti
Graduate Poster
- Joined
- Apr 4, 2023
- Messages
- 1,799
Race does not exist. But don't tell that to AI.
Study finds that artificial intelligence can determine race from medical images
Norman Alexander
Penultimate Amazing
That depends. When I was younger (MUCH younger), IQ tests were supposed to find where you fitted on some sort of standardised bell curve of population IQ. The problem was that those who did "best" in the tests turned out to be high fit because the curve itself was already calibrated against what were considered "success factors" for intelligence. We were really being fitted not to a general intelligence curve but to a "success in [our] society" curve.
As an extreme example, if the "IQ curve" was the result of assessing 10,000 Donald Trumps, and my results were high on that curve, I would have been assessed as "high IQ". Never mind that none of those Donald Trumps could spell their names correct two times out of three.
More generally, the assessment comes down to what is considered "high intelligence". And that is very much societal.
Indeed - socio-economic factors differentiating "IQ" results. So if you have a high socio-economic upbringing then you score better. As I said above, IQ depends on what you are using as a measure.
My example of Turkish is not about language. It's about differences in what is considered "intelligent". The factors for a white Anglo male who lives and works in a very technical western computing environment (i.e. me) are going to be very different to those for a Turkish archeologist or sociologist.
I hope I have made it clear that what an Amazonian considers "intelligent" is someone who finds newer, cleverer, surer ways to trap that tapir, not my server skills, macro programming or database optimisation. My 132 IQ will likely count for nothing when it comes to tapir trapping. They would consider me a stupid dumbass, low IQ, and rightly so.
You know the axiom "the cunning of the stupid"? We use it a lot when talking about grifters like Trump and EmptyG. By any measure, these are stupid people, low IQ. But they are westerners, English-speaking (well...to an extent
), and supposedly educated. And yet they are "successful" at what they do. Why is that?My answer is that what we consider "success", what we measure "success" by, is the problem. Many people think they are indeed "successful". They have reached the pinnacle of their mountains. People actually think they are smart, i.e. high IQ. You and I don't because we measure them by a different yardstick. To us, they are dumb as ****, thick as a brick, low IQ.
So yes, context, environment, matters for IQ.
Trausti
Graduate Poster
- Joined
- Apr 4, 2023
- Messages
- 1,799
Singapore, at independence in the 1960s, had about 70% poverty. Within a generation it's economy was among the best in the world. No natural resources. Only human capital. Average IQ 106. (See also, South Korea.)
Last edited:
arthwollipot
Observer of Phenomena, Pronouns: he/him
You shouldn't believe everything you hear.
Wikipedia said:
https://en.wikipedia.org/wiki/Stephen_Jay_Gould
Trausti
Graduate Poster
- Joined
- Apr 4, 2023
- Messages
- 1,799
arthwollipot
Observer of Phenomena, Pronouns: he/him
Non sequitur.
jt512
Illuminator
- Joined
- Sep 24, 2011
- Messages
- 4,303
Correlation is not defined for a single measure of two variables on one unit of observation (you, in this case). Whether your misuse of the word is due to a lack of verbal or mathematical ability is, in this case, curiously unclear.
That said, verbal and mathematical ability, as measured by standardized tests, are strongly positively correlated, suggesting that there is an underlying factor—a general intelligence—that contributes to both.
Last edited:
Norman Alexander
Penultimate Amazing
Rain Man.
Yeah, I'm fully aware that was a scripted movie for dramatic effect. But it stands as an example for where verbal and mathematical abilities do NOT correlate strongly together to confirm any "general intelligence". There is a spectrum from there to "absolute normal", where intelligence simply does not correlate reliably.
I have known and worked with 140-plus IQ people who are so frickin' maladroit they can't learn to boil an egg reliably and have no idea where their money goes. I have also worked with 90-minus IQ people who have come up with some rather clever ideas (in computing). There is no really strong correlation, just a vaguely reasonable probability that a normally intelligent person can learn to drive safely and balance their check book.
Last edited:
jt512
Illuminator
- Joined
- Sep 24, 2011
- Messages
- 4,303
All your post shows is that you, like arthwollipot, don't understand what correlation is.
The two analyses I have seen both measured the correlation between verbal and mathematical ability to be above 0.8.