OOTS #1259 - The Discussion Thread

[Peelee]

IQ tests

Personal opinion here, I don't find IQ tests to be worth the paper they're printed on.

[DaOldeWolf]

Personal opinion here, I don't find IQ tests to be worth the paper they're printed on.

Human intelligence is hard to measure in general.

[Liquor Box]

Personal opinion here, I don't find IQ tests to be worth the paper they're printed on.

Is there a different measure you prefer? Your posts on the last page suggest that you don't hold self assessment in high regard.



Seperately (and not directed at anyone in particular), when I agreed with the statement that intelligence was overrated, i was meaning in the context of selecting a partner. It can be very useful for earning a good living. I just don't think it's that high on the list of what most people look for in a sexual partner or relationship.

[Thermophille]

when I agreed with the statement that intelligence was overrated, i was meaning in the context of selecting a partner. It can be very useful for earning a good living. I just don't think it's that high on the list of what most people look for in a sexual partner or relationship.

I apologize if I gave the impression that intelligence is worthless, looking back, I think my point came across more generally than I meant it, since I thought the context was clear. If everyone had an intelligence score of 10, we'd still be in the dark ages, if that.

My point is that I don't think it has much in the way of personal value, which I realize is kind of a fuzzy concept, but I think that anyone who considers their intelligence a point of pride is two steps behind someone who isn't intelligent or prideful.

[danielxcutter]

Personal opinion here, I don't find IQ tests to be worth the paper they're printed on.

I think the inventor of them made them to figure out which kids needed special help when learning instead of them to be an objective measurement of intellect and competence... so of course that's exactly what everyone used it for.

[Thermophille]

I think the inventor of them made them to figure out which kids needed special help when learning instead of them to be an objective measurement of intellect and competence... so of course that's exactly what everyone used it for.

Wasn't there also some racism there?

[bunsen_h]

I don't intend to speak against the value of intelligence. It's important, and I find it frustrating to deal with people of low intelligence. But having spent too much of my life trying to cope with highly-intelligent cranky *******s, and having been a highly-intelligent bit-of-a-jerk myself, I now see more value than I used to in wisdom, honesty, and generosity of spirit. At least when it comes down to assessing the nebulous "what is the worth of this person as a person".

[danielxcutter]

Wasn't there also some racism there?

Frankly I don’t remember, but racists do tend to latch onto any flimsy justification they can find so… maybe?

[Fyraltari]

IQ tests are bunk. As was said above the original test (who wan't quite an IQ test but it's the basis) was aimed at finding children the school system was failing in order for them to receive lessons tailored to their needs.

The modern IQ tests are based on several unproven assumptions, chief among them the existence of "general factor" that governs every form of intelligence (again, never proven to exist) and the idea that this general factor follows a normal distrubution (mathematic term, it maps to a bell curve ie most peope are close to average with fewer outliers the farther you are from the average, think of the distribution of the results of 5d6) with no justification for why it should which means that the results of IQ testing are often heavily distorted to make it fit the expected results. Most damning over is the established fact that the more IQ tests you take, the higher you score, you can train for them, meaning they don't measure an inherent part of your nature but your ability to take the bloody tests.

There is indeed a history of using IQ testing to justify racist/classist/eugenist policies, by using studies that shows that certain ethnic or social groups perform worse than others and those tests. These studies always ignore thr context such as pre-existing racist policies (like the ****ing Apartheid) or worse access to education (often meaning poor literacy levels) or cultural assumptions within the questions or children taking tests in a language that isn't their native tongue that ensure poor results.

TL;DR: Idiocracy is not a documentary.

[Thermophille]

I'm not a statistician, but doesn't almost everything based on random variation follow a standard distribution?

[Fyraltari]

I'm not a statistician, but doesn't almost everything based on random variation follow a standard distribution?

Everything that is the result of several identical independent random events, yes. There is no evidence that differences in cognitive ability are caused by such.

Edit: and again, the fact is that even when large groups pass the same test, the results don't map to a bell curve and have to be artificially transformed to do so.

[mjasghar]

The problem with most of these tests is they end up being more about doing well at test than what they are testing for
Specifically with iq tests they assume high levels of education in formal English etc. Most importantly the non verbal reasoning logic tests are actually the worst as you can get higher results for them if you have practice questions so you know how they are structured. It’s why private schools in the uk do so well - they have their own private exam board so they only need past papers from that board for the pupils to memorise.
Unfortunately statisticians rarely see their own prejudices- I had a stats lecturer whose introduction to the subject was about the red car fallacy. Yet years later I heard he had made comments about race and iq having fallen for that fallacy himself.

[Liquor Box]

IQ tests are bunk. As was said above the original test (who wan't quite an IQ test but it's the basis) was aimed at finding children the school system was failing in order for them to receive lessons tailored to their needs.

The modern IQ tests are based on several unproven assumptions, chief among them the existence of "general factor" that governs every form of intelligence (again, never proven to exist) and the idea that this general factor follows a normal distrubution (mathematic term, it maps to a bell curve ie most peope are close to average with fewer outliers the farther you are from the average, think of the distribution of the results of 5d6) with no justification for why it should which means that the results of IQ testing are often heavily distorted to make it fit the expected results. Most damning over is the established fact that the more IQ tests you take, the higher you score, you can train for them, meaning they don't measure an inherent part of your nature but your ability to take the bloody tests.

If the assumption that human intelligence maps to a normal distribution is incorrect, that wouldn't undermine the idea that iq correlates with intelligence. Any distortion would not change that a better testing result would lead to a better iq score. I may mean that the extent of the difference in iq scores is not a good representation of testing results.

Your other reason you give is that g factor is not proven. That's true, it's probably unproveable. But observable data (eg. performance across a wide variety of different mental tasks correlating) is consistent with somethin of its nature existing.

We do know that iq score is a predicter to some degree of several things that we might intuitively think would be intelligence based (eg, income, school performance, job performance). It's possible that this might be because of some cofounding variable, but it has been replicated in multiple studies, which have controlled for things like economic status. So even if we can't prove why iq tests work, they do work to some extent.

It is of course widely acknowledged that iq does not capture all aspects of things which we might consider intelligence (I think creativity is something which it is widely accepted to not accurately capture.

So it ought to be taken for what it is - a system for measuring some aspects of human intelligence, which is not perfect, but does produce results that are meaningful for some purposes.

Which, if you thinks about it, puts it in the same place as any other testing. The tests people take to qualify as doctors, lawyers, engineers or other such professions also rely on tests that are imperfect (eg taking more tests makes you better at them), have limitations, but still produce useful results.

I suggest that there's a middle ground between saying that they're 'bunk' and them being a perfect representation of human intelligence.

[Peelee]

We do know that iq score is a predicter to some degree of several things that we might intuitively think would be intelligence based (eg, income

I'm going to press X to doubt that IQ tests have any sort of prediction ability towards income, which is wholly irrelevant to intelligence.

[Liquor Box]

I'm going to press X to doubt that IQ tests have any sort of prediction ability towards income, which is wholly irrelevant to intelligence.

I'm surprised that you think intelligence is wholly irrelevant to income. It seems to me intuitive that there would be some relationship between intelligence and income given that several high paying jobs (eg doctors) require decent intelligence. If a person had significantly below average intelligence that would seems to make a whole lot of profession choices unlikely to be obtained. More broadly lots and lots of roles (eg administrative roles) require some degree of education which gets easier the more intelligent you are. If nothing else, intelligence would offer an advantage in terms of figuring out how to make more money.

Absolutely not the only factor, quite likely not even the most important factor. But I can't accept that intelligence is entirely irrelevant to income.

Anyway, here are a couple of articles on the point of IQ predicting income. If you want actual studies, I'm sure i can find some
https://www.expensivity.com/iq-and-salary-connection/
https://slate.com/human-interest/201...s-in-life.html

[hroþila]

But if IQ tests are biased towards people with a particular kind of education that folks from existing high-income families are more likely to get, isn't the whole thing just kinda circular?

[Liquor Box]

But if IQ tests are biased towards people with a particular kind of education that folks from existing high-income families are more likely to get, isn't the whole thing just kinda circular?

As I mentioned a couple of posts back, there are studies which have controlled for economic circumstances (so have compared the results of people with similar economic backgrounds) and have still produced results.

That doesn't mean that that economic background doesn't influence the results somewhat. If so, that would be an imperfection in the validity of the score, but it suggests the score measure something more than that.


Edit:
I googled it and found this study.
https://medcraveonline.com/JNSK/fact...-quotient.html
It concludes that IQ is influenced by both environmental and inherent factors, and says at one point that socioeconomic status is one factor that correlates with IQ (if you want to find it, I suggest a wordsearch of the term 'economic').

[Peelee]

I'm surprised that you think intelligence is wholly irrelevant to income. It seems to me intuitive that there would be some relationship between intelligence and income given that several high paying jobs (eg doctors) require decent intelligence.

Such high paying jobs require decent schoolwork, not necessarily intelligence.

Further, from those very articles you posted:

Researchers have generally found that there is a positive correlation between IQ and salary.

Correlation does not equate to being a predictor. Correlation could be wholly unrelated. And hey, I wonder if one of the articles you posted says as much.

An IQ score correlates with success but doesn’t dictate it, and it is a poor “measure of the man.” Allow me to be politically incorrect for a moment to make my point: Saying that IQ scores predict success is a little like saying that the color of your skin at birth predicts your future income.

Oh my, it appears as if the very articles you put up as evidence that it is a predictor openly say "this is not a predictor".

[Liquor Box]

Such high paying jobs require decent schoolwork, not nevessarily intelligence.

True, but intelligence is at least a factor in being succesful in school. Again, absolutely not the only factor, but a factor.

IQ does predict success at school at least very strongly, so if you think school success is a predictor of income it stands to reason IQ would be too)

Further, from those very articles you posted:

Correlation does not equate to being a predictor. Correlation could be wholly unrelated. And hey, I wonder if one of the articles you posted days as much.

Oh my, it appears as if the very articles you put up as evidence that it is a predictor openly say "this is not a predictor".

Saying it does not dictate is different from saying it does not predict. Dictating it makes it certain, predicting it only means it is more likely.

As ot what the articles say about IQ predicting income:

If you take a “real” IQ test (see comments below), then the result is a strong statistical predictor of multiple future life outcomes—income, education level, health, even longevity.

https://slate.com/human-interest/201...s-in-life.html

It would not be inaccurate to say that a higher IQ can predict a higher salary.

https://www.expensivity.com/iq-and-salary-connection/

Here's what Wikipedia says on the point (usefully noting the results when controlling for parental socialeconomic status):

The American Psychological Association's 1995 report Intelligence: Knowns and Unknowns stated that IQ scores accounted for about a quarter of the social status variance and one-sixth of the income variance. Statistical controls for parental SES eliminate about a quarter of this predictive power.

[Peelee]

As ot what the articles say about IQ predicting income:

Let's look at the very next sentence that you notably didn't quote.

There are loads of studies that confirm these correlations. So in that sense, it “predicts” your future “success.” However—and it’s a big however—be careful what inferences you draw from this answer.

Oh my, with all those quotation marks it appears as if they are trying very much to distance themselves from this idea. And hey, all those things it "predicts", as luck would have it, those are already all predicated on having a safe, stable, well-off environment to start with. Why, one might wonder if the types of people who get formal IQ testing are already the exact demographic to be correlated with such things!

I hope Nick Cage stops making movies entirely, we can significantly reduce drownings by falling into a pool that way.

[Liquor Box]

Let's look at the very next sentence that you notably didn't quote.

Oh my, with all those quotation marks it appears as if they are trying vrtu much to distance themselves from this idea. And hey, all those things it "predicts",

Yep it goes on to say

Don’t equate IQ with intelligence. It’s still a matter of, at times heated, debate as to what IQ actually measures. And it’s also a complex conversation to discuss what we mean by intelligence.

Which funnily enough, I also acknowledged.

Like I said, it is a predictor of income, but it is not on all fours with what we think of as intelligence.

as luck would have it, those are already all predicated on having a safe, stable, well-off environment to start with. Why, one might wonder if the types of people who get formal IQ testing are already the exact demographic to be correlated with such things!

As noted in three of my previous posts, several studies have controlled with economic status. See the quote from the wikipedia article for example.

[Peelee]

Yep it goes on to say


Which funnily enough, I also acknowledged.

Like I said, it is a predictor of income, but it is not on all fours with what we think of as intelligence.

The article explicitly says it "predicts" in the sense that it correlates. Which is exactly as much evidence that Nicholas Cage movies predict falling-in-pool drownings. Which is to say, it doesn't.

[Liquor Box]

The article explicitly says it "predicts" in the sense that it correlates. Which is exactly as much evidence that Nicholas Cage movies predict falling-in-pool drownings. Which is to say, it doesn't.

No it doesn't. It says that it is a "strong statistical predictor of...".

Cage movies are not a strong statistical predictor of falling in a pool because Cage only appears in a small number of movies a year, not enough for it to be a statistical predictor - they are not dependent. Your link is the very reason why statisitic require sufficient numbers to yield statisitically significant results. If you don't you just get odd coincidences from independent numbers, like Cage.

[Peelee]

No it doesn't. It says that it is a "strong statistical predictor of...".

Cage movies are not a strong statistical predictor of falling in a pool because Cage only appears in a small number of movies a year, not enough for it to be a statistical predictor - they are not dependent. Your link is the very reason why statisitic require sufficient numbers to yield statisitically significant results. If you don't you just get odd coincidences from independent numbers, like Cage.

I find it the height of irony that the article explicitly says to be careful what inferences you draw from this, and you go on to draw the exact inferences it says should not be drawn.

Maybe another quote from that article might help.

Allow me to be politically incorrect for a moment to make my point: Saying that IQ scores predict success is a little like saying that the color of your skin at birth predicts your future income. You can see the problems: Causality is an issue, the significance of other factors is an issue, and so on

[Liquor Box]

I find it the height of irony that the article explicitly says to be careful what inferences you draw from this, and you go on to draw the exact inferences it says should not be drawn.

Maybe another quote from that article might help.

If you cannot see the problems as the author has and as I have, then might I suggest this could be the crux?

Oh I completely get that side of it. I wasn't suggesting any inference from it other than that it was a predictor. Ethnicity is of course a predictor of income in USA, and so is IQ. The crux was perhaps that you were answering a different point from the one I was making?

The studies which control for socioeconomic status, and the studies that talk about the different factors that influence IQ (quoted or referenced previously) might have more value if we want to draw inferences. Those suggest that both environmental factor (like socioeconomic factors) an non-environmental factors (like inherent intelligence) are part of it.

At the end of the day, it's neither meaningless nor perfect, in my opinion at least.

[Peelee]

At the end of the day, it's neither meaningless nor perfect, in my opinion at least.

The article ends by stating it is meaningless, I would like to point out.

[Liquor Box]

The article ends by stating it is meaningless, I would like to point out.

It ends by stating that it is a predictor, and then poses the question "so what?"

To restate - I accept the article says that we should not jump to social conclusions based on the fact that IQ is a predictor of income. I'm not suggesting any social conclusions either.

[Peelee]

It ends by stating that it is a predictor, and then poses the question "so what?"

To restate - I accept the article says that we should not jump to social conclusions based on the fact that IQ is a predictor of income. I'm not suggesting any social conclusions either.

The article states "yes, this is technically true, if you look at it from a very specific angle and squint and then don't make any assumptions based on it, so.. So what?" If in slightly longer and less blatant phrasing. The "so what?" is fairly telling. It makes no sense if it purports that it's in any way an accurate or useful metric. Because it isn't. As the article explains in detail. Using that article as evidence that it is an accurate or useful metric is certainly an interesting take.

[Liquor Box]

The article states "yes, this is technically true, if you look at it from a very specific angle and squint and then don't make any assumptions based on it, so.. So what?" If in slightly longer and less blatant phrasing. The "so what?" is fairly telling. It makes no sense if it purports that it's in any way an accurate or useful metric. Because it isn't. As the article explains in detail. Using that article as evidence that it is an accurate or useful metric is certainly an interesting take.

The article does not in fact contain the quote you ascribe to it. That might be your take from the article (and if so, I don't think it's accurate), but it is not something the article says and should not have been presented in quotations. I have only quoted what it actually says.

We may be at cross purposes as to what we are discussing here. I thought we were discussing you view that IQ was not a predictor of income because of your surprising belief intelligence is "wholly irrelevant" to income. I referenced the article for that point alone (it was the first that came up on a google search) and it is explicit in several places that IQ tests are "a strong statistical predictor of... income". The article nowhere implies that any degree of squinting is necessary to reach this conclusion, and even notes that "there are loads of studies" that confirm this. I note further that this is only one of three sources I quoted to support my claim that IQ is a predictor of income.

Are we now finished with the question of whether IQ is a predictor of income, or do you still doubt it?


If we are finished with that, I would be happy to discuss the seperate topic of whether IQ is accurate or useful. I refer you back to my original comments on the point. I said that IQ tests do not (and do not claim to) represent the full spectrum of what we think of as human intelligence, and are instead more limited. Even in that limited capacity they are not perfect, but do produce results that are meaningful for some purposes. What do you think? What do you think its shortcomings are?

I get the problem that there are covariables, like socioeconomic status, that make the results less accurate at predicting the underlying intelligence, so if that is your point I agree,

I must admit, I get the feeling from you that there's an undercurrent of this being more than an academic discussion about dry subject matter. Are you concerned about some implication arising from IQ tests having some meaning that I am missing?

[Peelee]

The article does not in fact contain the quote you ascribe to it.

I did not ascribe anything to th article except in quote blocks. Are you referring to my paraphrase, which I was very clear about being paraphrased?

We may be at cross purposes as to what we are discussing here. I thought we were discussing you view that IQ was not a predictor of income because of your surprising belief intelligence is "wholly irrelevant" to income. I referenced the article for that point alone (it was the first that came up on a google search) and it is explicit in several places that IQ tests are "a strong statistical predictor of... income". The article nowhere implies that any degree of squinting is necessary to reach this conclusion

Except where it does, which I ahve already quoted directly in previous replies. If you wish to ignore all the caveats given in the article that all but openly state that its simple correlation without know causation, feel free, but that's a path I will not follow you down.

Are we now finished with the question of whether IQ is a predictor of income, or do you still doubt it?

Of course I still doubt it, because nothing you have shown or said has given me any reason to believe it and, in fact, your own sources speak to how they are not useful for doing so.

That being said, I am perfectly content to be finished with the question of whether IQ is a predictor of income, since if you post articles that go against your own view and see them as supporting it, I doubt anything I can say would be of any use.