Topic: I.Q.

[arthwollipot]

There was a brief mention of I.Q. in this week's show, and it got me wondering. Is I.Q. actually a useful measure of anything? I was brought up to believe that it didn't really reflect intelligence (whatever that is) - all it did was measure someone's ability to do I.Q. tests.

Can "intelligence" even be defined in such a way as to be easily measurable with a standard test?

[Hanes]

Eh, it's good for measuring a person's ability to solve problems like those found on IQ tests.  Those questions tend to require a certain form of abstract reasoning.

It doesn't measure other important aspects of intelligence, such as creativity.

[Cowtown Cody]

IQ tests are based on the theory that there is one, unified General Intelligence, g factor that accounts for how smart somebody is, and IQ tests historically have been crafted to measure the strength of that factor.  Reading about the g factor will probably help explain a lot.

One major idea is that it doesn't matter what the puzzle or question you're asking is.  If it's truly General Intelligence, then it will affect your abilities no matter what question is put in front of you.  On high-quality intelligence tests, there's a lot of concurrence - an IQ test designed for people who cannot read, which necessitates different kinds of questions - tends to give a score that's very similar to standard IQ tests like the Stanford-Binet or the WAIS-R.

So it seems fair to say that the g factor, according to all the psychometric testing that's been done so far, might really exist.  It doesn't seem to go away when we change how we conduct the tests, and it seems to influence to varying degrees how good we are at solving different kinds of problems.  So for a basic definition of intelligence, for the purposes of psychometrics, we can use "the ability to solve problems of varying nature and complexity."

Which doesn't mean we know exactly what the g factor is, or by what mechanism it works, or anything like that.  We're only observing on a very wide array of tests what appear to be the output of a single factor of intelligence, which we're calling "g factor."  There are other theories out there, but this is the one that is the current consensus.

IQ does have some predictive validity in a range of phenomena like academic success, job training success, success on other tests like the SAT, as well as a handful of others.  But they're easy to overstate.

[Trinoc]

It may be a logical fallacy on my part, but whenever the subject of IQ comes up I remind myself that the concept was devised in a set of biased (to put it mildly) observations intended to show that intelligence is inherited, and so support the idea of racial superiority espoused by a certain movement at the time which Godwin's Law prohibits me from naming.

Most of the ideas about IQ since then seem to have been built on these dodgy foundations, so we need to ask: if we could magically remove the work of people like Cyril Burt from the history of the subject, would any of the subsequent stuff continue to stand, or would it collapse in a heap of pseudoscience?

[Cowtown Cody]

Cyril Burt removed, g factor would still be something that is taken very seriously in psychometric testing.  Even when you strip all the racists and eugenicists out, g is still there in the statistical analyses and its presence is uncontroversial.  The statistics say it's there.  This factor hasn't gone away with time or repetition. 

Modern, high-quality intelligence tests built and normed by scientists who are keenly aware of IQ testing's dark roots demonstrate the g factor as well.  Those norms are matched to the test taker so they know how you did compared with people your same age, gender, race, ethnicity, socioeconomic background, educational background, etc.  So a lot of the old, really bad problems are much harder to have now.  Believe it or not, we are better scientists now than we were in 1909.  It's not to say that there aren't still problems, but I do mean to say that comparing a top of the line, modern intelligence test to the early work of the eugenicist set is a lot like comparing a supercar to a turnip truck.

The question now is one of how the factor is derived - essentially, if the g factor is one thing that operates across many types of problems, or if it's actually many different types of specific intelligence modules, each operating only on one specific type of problem, thereby explaining why the g factor seems to be more active on some tests and in some areas rather than others.  How you get there instead of to a unified g factor involves a lot of very advanced statistics and mathematics but it can be done. 

A good intelligence test now, like the Stanford-Binet will also provide subscales so you can see where you were comparatively better or worse, in some general categories like spatial reasoning, verbal reasoning, numerical reasoning, etc.  Whether intelligence is one thing or many modules, you can still see and compare your abilities across types of problem, so the theoretical issue doesn't actually matter much at the point where the tests contact the public.

The other thing to remember is that IQ testing like all psychometric testing is an indirect measure, and that makes it weaker but we have to be careful not to heap all indirect measurements into the "pseudoscience" category.  It cannot access what is truly in your brain, and it is forced to rely on repeatability, concurrence, and prediction for its validity.  Modern intelligence tests - good ones - agree with each other pretty closely about how they score the same person, how the same person does over time, and they generalize and predict results in other fields fairly well - not fantastically, but say, moderate to high-moderate depending on what you want to look at.

Modern, high quality intelligence tests like the WAIS-R and the Stanford Binet are completely different animals from their predecessors, and we're much better off for it.

[Shadow Of A Doubt]

If I'm remembering correctly, IQ scores stay relatively constant compared with peers from the ages of 5-6 to adulthood which also tends to suggest that it is a proxy for something and it doesn't seem out of line to call that thing general intelligence.

But it is often used by idiots in an attempt to bolster an argument from incredulity or authority, and in that scenario it's patently a red herring.

[andrewclunn]

IQ correlates with income, so there's probably something to it.  I recall a video that panda posted a while back where a black British journalist is investigating if there is a difference between the races.  He takes an IQ test, and when he bombs horribly he starts talking about how useless they are.

[vociferous]

IQ correlates with income, so there's probably something to it.  I recall a video that panda posted a while back where a black British journalist is investigating if there is a difference between the races.  He takes an IQ test, and when he bombs horribly he starts talking about how useless they are.

Yes, there is an interesting component to that which is researched but rarely discussed in the press.

IQ tests have consistently shown that blacks score 10-20 points lower than average, no matter the country or continent.  This has been a classic nature versus nurture question and the source of much research.

The nurture proponents point to evidence such as Native American scores, which were dismally low a few decades ago but now are about average.  The nature proponents point to the fact that the gap is persistent, no matter the place: North America, South America, Africa, Europe. . . 

It is an open and politically sensitive question. 

[Hanes]

IQ correlates with income, so there's probably something to it.  I recall a video that panda posted a while back where a black British journalist is investigating if there is a difference between the races.  He takes an IQ test, and when he bombs horribly he starts talking about how useless they are.

Yes, there is an interesting component to that which is researched but rarely discussed in the press.

IQ tests have consistently shown that blacks score 10-20 points lower than average, no matter the country or continent.  This has been a classic nature versus nurture question and the source of much research.

The nurture proponents point to evidence such as Native American scores, which were dismally low a few decades ago but now are about average.  The nature proponents point to the fact that the gap is persistent, no matter the place: North America, South America, Africa, Europe. . . 

It is an open and politically sensitive question.

Quick google search shows that there are studies that confirm my suspicion:
Northwestern study finds that poverty and early learning opportunities -- not race -- account for the gap in IQ scores between blacks and whites.

As in many other studies, the black children in the study had IQ scores a full 15 points lower than their white counterparts. Poverty alone,the researchers found, accounted for 52 percent of that difference, cutting it to 7 points. Controlling for the children's home environment reduced the difference by another 28 percent, to a statistically insignificant 3 points -- in essence, eliminating the gap altogether.

I'm skeptical because, as I said, it confirms my prior suspicion.  But the alternative is so implausible on the face of it: a few thousand years of evolution, with essentially the exact same environment (as far as mental abilities are concerned), could be expected to produce stark differences in mental abilities?  My skepticism of that far outweighs that of the alternative.

[vociferous]

IQ correlates with income, so there's probably something to it.  I recall a video that panda posted a while back where a black British journalist is investigating if there is a difference between the races.  He takes an IQ test, and when he bombs horribly he starts talking about how useless they are.

Yes, there is an interesting component to that which is researched but rarely discussed in the press.

IQ tests have consistently shown that blacks score 10-20 points lower than average, no matter the country or continent.  This has been a classic nature versus nurture question and the source of much research.

The nurture proponents point to evidence such as Native American scores, which were dismally low a few decades ago but now are about average.  The nature proponents point to the fact that the gap is persistent, no matter the place: North America, South America, Africa, Europe. . . 

It is an open and politically sensitive question.

Quick google search shows that there are studies that confirm my suspicion:
Northwestern study finds that poverty and early learning opportunities -- not race -- account for the gap in IQ scores between blacks and whites.

As in many other studies, the black children in the study had IQ scores a full 15 points lower than their white counterparts. Poverty alone,the researchers found, accounted for 52 percent of that difference, cutting it to 7 points. Controlling for the children's home environment reduced the difference by another 28 percent, to a statistically insignificant 3 points -- in essence, eliminating the gap altogether.

I'm skeptical because, as I said, it confirms my prior suspicion.  But the alternative is so implausible on the face of it: a few thousand years of evolution, with essentially the exact same environment (as far as mental abilities are concerned), could be expected to produce stark differences in mental abilities?  My skepticism of that far outweighs that of the alternative.

The long and short of it is, at least from my perspective as someone trained in the quantitative sciences and not the social sciences where speculative reasoning is more permitted, there just is not much good evidence to support either hypothesis.  We know that a significant amount of intelligence is genetically linked (which would seem to support the nature hypothesis) but we cannot rule out possible environmental effects, especially prenatal and early childhood on IQ (but at the same time, that is undermined by the persistent gap across countries and continents).

Bottom line is, it is an interesting question, but until we understand more about the physical and genetic basis of intelligence (I believe neurologists and psychologists call it g), it just is an unanswerable question. 

[andrewclunn]

Does IQ test a particular type of reasoning that (while not being the only measurement of intelligence) does measure a particular type of reasoning and mental ability?  Yes.

Is there a positive correlation between IQ test scores and financial success (in the form of increased income)?  Yes.

Is there a genetic component to IQ (saying that it is in part based on inheritance)?  Yes.

Is there a correlation between classical racial categories and certain genetic variations that impact IQ?  Danger WIll Robinson!  Danger!

[Haricots]

I'm skeptical because, as I said, it confirms my prior suspicion.  But the alternative is so implausible on the face of it: a few thousand years of evolution, with essentially the exact same environment (as far as mental abilities are concerned), could be expected to produce stark differences in mental abilities?  My skepticism of that far outweighs that of the alternative.

Differing selective pressures causing changes in gene frequencies is not the only way differences between populations can arise. Other obvious explanations would be the founder effect, preexisting genetic diversity in Africa and hybridization.
I'm still skeptical, but I can't see any good reason to definitely rule it out as extremely improbable.
The Flynn effect, the observation that IQ seems to be increasing in populations, is what makes it hard for me to not suspect non-genetic causes of IQ differences between populations.

We know that a significant amount of intelligence is genetically linked (which would seem to support the nature hypothesis)

I'd say that it is more of a prerequisite for racial differences than a good argument for it.
What is important to note is that these studies study the degree of influence genes have in the population that is studied.
This implies that one would observe a stronger and stronger degree of genetic influence as the diversity in factors that influence IQ that are unrelated to genes (or at least less than IQ) is decreased.
So I would get a much lower number if I for some reason decided to do the study on 50% Americans and 50% hunter gatherers than if I did a study on only one of those populations.
My point is that studies on one population doesn't translate well to questions about differences between populations (or a population at different times for that matter, so we shouldn't suspect that the Flynn effect is caused by smart people getting laid more because of these studies  )

[vociferous]

We know that a significant amount of intelligence is genetically linked (which would seem to support the nature hypothesis)

I'd say that it is more of a prerequisite for racial differences than a good argument for it.
What is important to note is that these studies study the degree of influence genes have in the population that is studied.
This implies that one would observe a stronger and stronger degree of genetic influence as the diversity in factors that influence IQ that are unrelated to genes (or at least less than IQ) is decreased.
So I would get a much lower number if I for some reason decided to do the study on 50% Americans and 50% hunter gatherers than if I did a study on only one of those populations.
My point is that studies on one population doesn't translate well to questions about differences between populations (or a population at different times for that matter, so we shouldn't suspect that the Flynn effect is caused by smart people getting laid more because of these studies  )

True, but different selective pressures cause different groups to have different frequencies of a number of genes and we do know that IQ has a very strong genetic component.  It is also widely believed that IQ is something that is largely congenital and cannot be significantly increased past early childhood, which of course leaves a lot of room for early environmental effects, such as in the womb or in pre-adolescence. 

The biggest problem with this whole debate is you are starting with something that is empirically shown beyond any reasonable doubt to be true (the large, cross-cultural, below average IQ of blacks) and then trying to create ad hoc theories to explain it based primarily on a priori reasoning with virtually no empirical support, which makes all these theories extremely weak.