IQ Tests

I have never been convinced that IQ tests have really distinguished core intelligence from education.  I scored much better on IQ tests after I practiced and read about how to tackle certain types of problem.

It is for this reason that I have always assumed the Flynn effect to be due to education, not changes in native intelligence.

  • http://www.facebook.com/logan.durand Logan Durand

    Long time fan, Warren, but I think you've got some reading to do here.

    "I have never been convinced that IQ tests have really distinguished core intelligence from education."

    This is difficult for individual tests, but in the aggregate, they are very well suited to this purpose. This problem has been anticipated by intelligence researchers, and the tests are constructed in such a way as enable "true intelligence" (to whatever extent such a thing exists) to be teased out. A test battery will typically contain different sections, and different styles of questions. Through factor analysis, it is possible to derive the extent to which different components 'probe' an individual (or population's) 'real' intelligence, known as 'the g factor'. The degree to which a test proxies for g is referred to as its 'g-loading'. (That this 'real' intelligence is biologically real, and not merely a statistical artifact, has been established through other research.)

    "I scored much better on IQ tests after I practiced and read about how to tackle certain types of problem."

    From the available literature, it has been demonstrated that the 'trainability' of a test or test components is inversely related to that component's g-loading, so it is possible to compensate for such problems by weighing components accordingly. More broadly, though, it should come as no surprise that any testing device can be defeated by an aware subject. A child who wishes to get out of going to school knows how to manipulate a thermometer to exaggerate his fever. This does not mean the thermometer is a poor instrument for measuring 'real' temperature. Rather, it means the results need to be interpreted properly.

    "It is for this reason that I have always assumed the Flynn effect to be due to education, not changes in native intelligence."

    This is unlikely to be exclusive cause, as the phenomenon appears to be uniform across all age groups, even infants. Flynn also contends that the increase in nominal 'IQ' has been accompanied by an increase in g. (However, this correlation has been disputed by Rushton and Jensen 2010, so it's possible there's been some of both.)

    The Flynn Effect has also been observed to be concentrated in the lower end of the bell curve. The most plausible explanation appears to be improvements in nutrition and the womb environment. I suspect that dramatically improved access to good food in the industrialized world has made it possible for those with below-average genes for intelligence to make it a lot farther today than they would have yesterday.

    In either case, one thing appears to be clear - for first world, white populations, the Flynn effect has stopped for the past decade or so. Flynn supposes, and I would agree, that our environmental improvements have reached the upper limit of what we can do with our current genes. Even more depressing, our dysgenic fertility pattern is slowly eroding our average genetic potential for intelligence. I am cautiously excited to see what data the next decade will yield.

  • obloodyhell

    Now if only someone would come up with a Wisdom Quotient, we could limit the number of fools surrounding us. :-P

    While "Avoid Democrats" is a good rule of thumb, it seems insufficient with the riding number of RINOs.

    LOLOLOLOLOL...

  • http://www.facebook.com/matthew.slyfield Matthew Slyfield

    We don't even have a decent definition of intelligence that is wholly distinct from education. Claiming that this undefined thing can be reasonably measured is absurd

  • Andrew

    Everything I learned in getting a BA in sociology has convinced me we don't know how to define intelligence apart from culture and education.

    The "(to whatever extent such a thing exists)" part is the important part to me. Wake me up when we can clearly define what we're testing for, and then I'll start paying attention. Until then IQ testing is more sociopolitical than scientific in my opinion.

    (I'm fully aware of the hypocrisy inherent in a sociologist calling out anything or anyone for being "unscientific" :-P)

  • obloodyhell

    Ummm, the notion that we can't measure what we're measuring with an IQ test is inherently ludicrous. Whether it means anything is acked as potentially of debate, but the idea that we can't measure the ability to manipulate concepts and ideas in the general, as opposed to specific, sense is ridiculous.

    Does education and culture affect this? Of course it does. You have to have the notion of manipulating things to be able to do it at all, and that's cultural. But once the idea is there, yeah, some people have natural talent at it and others not so.

    I'm exceptionally good at math, I am a natural at it. Others have a hard time grasping the idea of manipulating abstract concepts in terms of "a"s and "b"s. To suggest that is entirely culture or education is absurd. It's rather blatant that some people have more ability in that regard than others. Can you learn to get better? Of course. Will you ever be better, with practice, than someone who is a natural and has had the same amount of practice? Eph no.

    There are other factors dealing with social interactions which one can have a natural talent with, while others have trouble with it. Politicians clearly are good with playing social games. Can you learn to get better? Of course. Will you ever be better, with practice, than someone who is a natural and has had the same amount of practice? Eph no.

    To claim otherwise is to say that, if I'd started early enough, I could have been as good a wide receiver as Jerry Rice. It's a preposterous notion on its face. Some people have natural talents that others don't. That's not even a measure of "better" as a person, it's just a measure of better for a given task. The only way you can argue "if you started early enough" is if you mean a few hundred thousand years ago... :-/

    An IQ test measures how good you are for a given set of tasks, most of which are particularly useful as a generalized functional member of society. Period.

  • http://www.facebook.com/logan.durand Logan Durand

    I added that phrasing because it is unclear if intelligence is a singular 'ball' of a concept or a set of tightly interconnected abilities that are mutually reinforcing. In either case, through factor analysis it is clear that all mental abilities load on a single common axis of abstract reasoning capability that is culture independent and highly heritable, with equal predictive power for all groups.

  • http://www.facebook.com/matthew.slyfield Matthew Slyfield

    "unclear if intelligence is a singular 'ball' of a concept or a set of
    tightly interconnected abilities that are mutually reinforcing"

    In other words exactly as Andrew said the people writing the IQ tests do not have a solid understanding of what they are trying to measure.

  • http://www.facebook.com/matthew.slyfield Matthew Slyfield

    The problem with IQ tests is that we don't really understand exactly what it is we are measuring.

  • Russ R

    My only observation is that those on the left seem to absolutely despise the concept of IQ (an individual's cognitive ability), and will grasp at almost any excuse to deny: 1) its very existence, 2) its predictive value, 3) its heritability, and 4) its significant and well-documented group differences.

  • obloodyhell

    No, you're confusing the fact that there is a term "intelligence" which is different from what IQ tests measure, which is a subset of "intelligence". There've been any number of studies showing and defining different aspects of "intelligence", of which IQ is only one.

    Wisdom, "Social Intelligence", etc., these are not measured by IQ tests yet are clearly aspects of the much more general concept of intelligence.

    IQ tests measure what they are intended to measure, which I believe would best be described as the capacity to learn from books and stored knowledge,

  • obloodyhell

    I dunno, dude, a lot of the above commenters have some issues figuring it out, too.

  • obloodyhell

    In science, measuring comes BEFORE understanding. Not after.

    http://xkcd.com/397/

    The Curies didn't understand what they were measuring, either.

    People ascribe the wrong qualities to IQ tests, this isn't really a debate about that. But that they do measure a consistent quality with some flexibility over the individual's lifespan isn't really debatable.

    Debate the meaning, yes, Debate that it's measuring an actual quality, no. This isn't a mental Orgone meter.

  • http://www.facebook.com/matthew.slyfield Matthew Slyfield

    If you don't understand what you are measuring no meaningful conclusions can be drawn from the measurements.

  • http://www.facebook.com/matthew.slyfield Matthew Slyfield

    At best they are measuring what you have already learned not the capacity to learn.

  • marque2

    seems like a variation of the Lake Wobegon effect

    http://en.wikipedia.org/wiki/Illusory_superiority

  • nehemiah

    Hypothesis comes before measuring and hypothesis generally results from a mixture of curiosity and understanding. My first day at the Naval Security facility I was stationed at, I was told to go around the building and record the frequency settings of all the radios, about 300 units. When I was done i was told to add up the frequencies and divide the total by the number of radios. I started to add them up (no adding machine or computers in that day). But then realized that my leg was being pulled since I understood that the average frequency that was being monitored had no meaning whatsoever.

  • Craig Howard

    But Flynn argues that modern TV shows and other entertainment can be cognitively demanding, and video games like those of the Grand Theft Auto series probably require more thought than solitaire.

    Oh, what nonsense. Intelligence is the ability to leave one's prejudices aside and analyze a situation for what it is. It is the talent for seeing reality for what it is.
    And, yes, though intelligence may be innate, it can also be taught.

  • marque2

    It is interesting that intelligence tests revolve around logic puzzles, and you can be trained to do well in logic puzzles. Just as the SAT is suppose to measure your intelligence, and yet you can take courses for the SAT and improve by 100 points easy. It is funny I took a course for the GRE and they explained the analogies - and it was like - Oh that is all they want - my GRE - in comparable SAT sections was 450 points higher - they say for most folks the difference is no more than 50 points.

  • http://www.facebook.com/logan.durand Logan Durand

    To be sure, some logic puzzles lose their g-loadedness over time, due to test familiarity, retesting, and training (te Nijenhuis et al., 2007), necessitating tests be re-normed from time to time. However, it remains the case that some test components are very robust and allow for intelligence to be reliably measured in a test and culture independent manner (Jensen 1998).

  • obloodyhell

    Again, ridiculous misunderstanding of the scientific process.

    Given the choice between a dozen geniuses in the lab or a couple of idiots
    who can do [good] field work, take the idiots. It is more important to observe
    the facts accurately than to be able to interpret them, because if you
    observe enough of them, they'll explain themselves.

    - David Gerrold -

    Certainly as one observes the facts -- collects data -- one gains the capacity to begin making inferences and then forming hypotheses, and from that devising experiments to test those hypotheses.

    But it starts out with you knowing jack and sh**. ALWAYS.

    }}} But then realized that my leg was being pulled since I understood that the average frequency that was being monitored had no meaning whatsoever.
    Yeah, hunting snark.

    Yes, the data one measures CAN be meaningless.

    Is anyone here actually claiming that IQ tests are MEANINGLESS? LOL. If so, then, well, we have nothing to discuss.

    Arguing about interpretation and causes for variances in score across a population is one thing... but claiming that there is nothing there being measured is just flat out ludicrous. There are significant correlations between IQ and success by "western standards of success". And enough evidence that that's a causal connection and not a coincidental one.

  • obloodyhell

    Ridiculous. A game of Jeopardy tests what you've already learned. A final exam tests what you have already learned. An IQ test more than amply connects with HOW quickly you can learn certain types of information, and that directly applies to capacity to learn. I would grant it also tests for how good your grasp of certain basic cultural memes are, and this certainly affects your capacity to learn that type of information, which is clearly relevant to why one might want to see the results of an IQ test.

    The notion that an IQ test can be tied to certain key cultural concepts I won't argue with. But claiming that it doesn't measure what it is DESIGNED to measure is just absurd.

    a) By the arguments being presented, I could take ANYONE classed as a "moron" (yeah, I know it's not used any more) and turn them into a genius with "just more teaching". In actual fact, while one can improve upon your IQ score by practicing certain skills, this does NOT reject the hypothesis that you've actually improved your overall capacity to learn things in general faster than before...

    b) By the arguments being presented, I could take someone culturally illiterate (i.e., say, *ANY* Australian aborigine who has remained in the outback with no knowledge of "our" western memes, and teach them to be a genius, by western standards. This is probably true for some individuals. My bet is that it'd be true in the end for about the same percentage of westerners who are geniuses, and the more of the abos you trained in this way, the closer it would resemble a standard IQ bell curve, with a minor but not extreme shift-and-transform applied to it.

  • http://www.facebook.com/logan.durand Logan Durand

    "In other words exactly as Andrew said the people writing the IQ tests do
    not have a solid understanding of what they are trying to measure."

    This statement is a non sequitur. The lack of a precise understanding for the inner workings of the general intelligence factor g does not support Andrew's statement that we cannot delineate or measure such an underlying ability independent of the educational and cultural context in which it is manifested. I added the qualification to my statement because, although future research may discover different facets to g, at present such 'multiple intelligences' remain unproven.

    For an analogy, even though I have no idea what gravity really is or how it works, its existence as an independent force affecting objects in a variety of contexts can be reasonably inferred from statistics and observation.

  • http://www.facebook.com/logan.durand Logan Durand

    "We don't even have a decent definition of intelligence that is wholly distinct from education"

    Not sure how someone can even say this today. From "Mainstream Science On Intelligence," 1994:

    "Intelligence is a very general mental capability that, among other things, involves the ability to reason, plan, solve problems, think abstractly, comprehend complex ideas, learn quickly and learn from experience. ... Intelligence, so defined, can be measured, and intelligence tests measure it well. They are among the most accurate (in technical terms, reliable and valid) of all psychological tests and assessments. They do not measure creativity, character, personality, or other important differences among individuals, nor are they intended to."

    This editorial was written by LInda Gottfredson and signed by 52 researchers in the field of intelligence, specifically to educate the public and dispel such nonsensical statements as yours. Not only can we reasonably measure intelligence, but we have a rough understanding of its biological correlates, such as brain size, myelination of neural pathways, and reaction times, all of which are known to have a large heritable component and cannot be reasonably said to be due to 'education' or any other talisman of the environment-only theorists.

    "At best they are measuring what you have already learned not the capacity to learn."

    Again, this is a laughable statement that stopped being taken seriously at least two decades ago. Whatever underlying ability or set of abilities are measured by IQ tests, it is clear that they impact learning ability and future educational success. If we maintained such agnosticism on intelligence as it is you propose, many scientific breakthroughs would have been impossible, such as understanding the negative impact of alcohol during pregnancy, the negative effects of inbreeding, or the importance of breastfeeding during early development, all of which clearly effect 'intelligence' independent of education to some non-zero degree.

    Whatever IQ measures, it matters, and you probably want more of it.

  • Craig Howard

    This is a measurement problem. The only lack of intelligence lies with those who believe that people are more intelligent now than previously.

  • Harry

    My consulting firm used to give the Iowa test to everybody, including secretaries, and we had a cutoff number that screened applicants. I know you cannot do that anymore.

    We needed people who could reason. If we had been a pro football team, we would have tested how they ran the forty, and whether they could count hut, hut.

    We always took a second look at borderline cases, assuming they showed up on time to take the test and were not drooling on their jeans.