May 3, 2009

Geoffrey Miller on IQ

From Spent:
The irony about general intelligence is that ordinary folks of average intelligence recognize its variance across people, its generality across domains, and its importance in life. Yet educated elites meanwhile often remain implacably opposed to the very concept of general intelligence, and deny its variance, generality, and importance. Professors and students at elite universities are especially prone to this pseudohumility. They socialize only with other people of extraordinarily high intelligence, so the width of the whole bell curve lies outside their frame of reference. I have met theoretical physicists who claimed that any human could understand superstring theory and quantum mechanics if only he or she was given the right educational opportunities. Of course, such scientists talk only with other physicists with IQs above 140, and seem to forget that their janitors, barbers, and car mechanics are in fact real humans too, so they can rest comfortably in the envy-deflecting delusion that there are no significant differences in general intelligence.

Even within my own field, evolutionary psychologists tend to misunderstand general intelligence as a psychological adaptation in its own right, often misconstruing it as a specific mental organ, module, brain area, or faculty. However, it is not viewed that way by most intelligence researchers who, instead, regard general intelligence as an individual-differences construct—like the constructs “health,” “beauty,” or “status.” Health is not a bodily organ; it is an abstract construct or “latent variable” that emerges when one statistically analyzes the functional efficiencies of many different organs. Because good genes, diet, and exercise tend to produce good hearts, lungs, and antibodies, the vital efficiencies of circulatory, pulmonary, and immune systems tend to positively correlate, yielding a general “health” factor. Likewise, beauty is not a single sexual ornament like a peacock’s tail; it is a latent variable that emerges when one analyzes the attractiveness of many different sexual ornaments throughout the face and body (such as eyes, lips, skin, hair, chest, buttocks, and legs, plus general skin quality, hair condition, muscle tone, and optimal amount and distribution of fat). Similarly, general intelligence is not a mental organ, but a latent variable that emerges when one analyzes the functional efficiencies of many different mental organs (such as memory, language ability, social perceptiveness, speed at learning practical skills, and musical aptitude). ...

In the 1970s, critics of intelligence research such as Leon Kamin and Stephen Jay Gould wrote many diatribes insisting that general intelligence had none of these correlations with other biological traits such as height, physical health, mental health, brain size, or nerve conduction speed. Mountains of research since then have shown that they were wrong, and today general intelligence dwells comfortably at the center of a whole web of empirical associations stretching from genetics through neuroscience to creativity research. Still, the anti-intelligence dogma continues unabated, and a conspicuous contempt for IQ remains, among the liberal elite, a fashionable indicator of one’s agreeableness and openness.

Yet this overt contempt for the concept of intelligence has never undermined our universal worship of the intelligence-based meritocracy that drives capitalist educational and occupational aspirations. All parents glow with pride when their children score well on standardized tests, get into elite universities that require high test scores, and pursue careers that require elite university degrees. The anti-intelligence dogma has not deterred liberal elites from sulking and ranting about the embarrassing stupidity of certain politicians, the inhumanity of inflicting capital punishment on murderers with subnormal IQs, or the IQ-harming effects of lead paint or prenatal alcoholism. Whenever policy issues are important enough, we turn to the concept of general intelligence as a crucial explanatory variable or measure of cognitive health, despite our Gould-tutored discomfort with the idea.

You’ve probably heard that IQ tests are now widely considered outdated, biased, and useless, and that there’s more to cognitive ability than general intelligence—there are also traits like social intelligence, practical intelligence, emotional intelligence, creativity, and wisdom. Strikingly, these claims originate mostly from psychology professors at Harvard and Yale. Harvard is home to Howard Gardner, advocate of eight “multiple intelligences” (linguistic, logical-mathematical, spatial, musical, bodily-kinesthetic, interpersonal, intrapersonal, and naturalist). Yale is home to Peter Salovey, advocate of emotional intelligence, and was, until recently, home to Robert Sternberg, advocate of three intelligences (academic, social, and practical). (To be fair, I think the notions of interpersonal, social, and emotional intelligence do have some merit, but they seem more like socially desired combinations of general intelligence, agreeableness, conscientiousness, and/or extraversion, than distinctive dimensions that extend beyond the Central Six.)

Is it an accident that researchers at the most expensive, elite, IQ-screening universities tend to be most skeptical of IQ tests? I think not. Universities offer a costly, slow, unreliable intelligence-indicating product that competes directly with cheap, fast, more-reliable IQ tests. They are now in the business of educational credentialism. Harvard and Yale sell nicely printed sheets of paper called degrees that cost about $160,000 ($40,000 for tuition, room, board, and books per year for four years). To obtain the degree, one must demonstrate a decent level of conscientiousness, emotional stability, and openness in one’s coursework, but above all, one must have the intelligence to get admitted, based on SAT scores and high school grades. Thus, the Harvard degree is basically an IQ guarantee.

Elite universities do not want to be undercut by competitors. They do not want their expensive IQ-warranties to suffer competition from cheap, fast IQ tests, which would commodify the intelligence-display market and drive down costs. Therefore, elite universities have a hypocritical, love-hate relationship with intelligence tests. They use the IQ-type tests (such as the SAT) to select students, to ensure that their IQ-warranties have validity and credibility. Yet, they seem to agree with the claim by Educational Testing Service that the SAT is not an IQ test, and they vehemently deny that their degrees could be replaced by IQ tests in the competition for social status, sexual attractiveness, and employment. Alumni of such schools also work very hard to maintain the social norm that, in casual conversation, it is acceptable to mention where one went to college, but not to mention one’s SAT or IQ scores. If I say on a second date that “the sugar maples in Harvard Yard were so beautiful every fall term,” I am basically saying “my SAT scores were sufficiently high (roughly 720 out of 800) that I could get admitted, so my IQ is above 135, and I had sufficient conscientiousness, emotional stability, and intellectual openness to pass my classes. Plus, I can recognize a tree.” The information content is the same, but while the former sounds poetic, the latter sounds boorish.

There are vested interests at work here, including not just the universities but the testing services. The most important U.S. intelligence-testing institution is the Educational Testing Service (ETS), which administers the SAT, LSAT, MCAT, and GRE tests. ETS is a private organization with about 2,500 employees, including 250 Ph.D.s. It apparently functions as an unregulated monopoly, accountable only to its Board of Trustees. Although nominally dedicated to the highest standards of test validity, ETS is also under intense legal pressure to create tests that “are free of racial, ethnic, gender, socioeconomic, and other forms of bias.” This means, in practice, that ETS must attempt the impossible. It must develop tests that accurately predict university performance by assessing general intelligence, since general intelligence remains by far the best predictor of academic achievement. Yet, since intelligence testing remains such a politically incendiary topic in the United States, it is crucial for ETS to take the position that its “aptitude” and “achievement” tests are not tests of general intelligence. Further, its tests must avoid charges of bias by yielding precisely equal distributions of scores across different ethnic groups, sexes, and classes—even when those groups do have somewhat different distributions of general intelligence. So, the more accurate the tests are as indexes of general intelligence, the more biased they look across groups, and the more flack ETS gets from political activists. On the other hand, the more equal the test outcomes are across groups, the less accurate the tests are as indexes of general intelligence, the less well they predict university performance, and the more flack ETS gets from universities trying to select the best students. ETS may be doing the best it can, given the hypocrisies, taboos, and legal constraints of the American cognitive meritocracy. However, it may be useful for outsiders to understand its role in higher education not just as a gate keeper but as a flack absorber [should be "flak catcher"]. ETS throws itself on the hand grenade of the IQ test controversy to protect its platoon mates (elite universities) from the shrapnel.

My published articles are archived at iSteve.com -- Steve Sailer

72 comments:

Anonymous said...

Where have I read similar critiques before? I hope Miller's book gives some credit where it's due!

Anonymous said...

Wow!

Kudos to Miller!

George Will has written two recent op-ed pieces addressing the legal issues that explain why cognitive tests aren't used by employers -- you guessed it, lawsuits due to group differences in outcome.

Basically, if the test produces group differences, the burden of proof is on the employers to prove the test is essential and necessary for the job, as though general intelligence isn't preferrable for nearly any job.

Thus employers have abandoned the tests, and instead use a college diploma as proof of ability. This has also contributed to the inflation of college tuitions.

Anonymous said...

All I can say is "Wow!"

IS Geoffrey Miller's book going to get released if it is that politically incorrect? Or am I missing something that immunizes him from PC attacks?

Melykin said...

I have met theoretical physicists who claimed that any human could understand superstring theory and quantum mechanics if only he or she was given the right educational opportunities.
---------------------------

This is hard to believe. Didn't these physicists ever work as TA's in first year courses? Didn't they mark first year exams? But I suppose if all their education and careers had been spent at elite universities, they might not understand what many first-year students are like in most colleges and universities. I teach math at a small university. No need to convince me or any of my colleagues that a lot of students have trouble with math.

Bob said...

The idea that very selective schools are just legal ways to give potential employees IQ tests is such BS.

Anyone who could get into Harvard could also go to a state school or a solid private school for free, including living expenses, and then list their 1600 SAT score on their resume (and have a certified copy handy for employers). Yet something like 80% of people admitted to Harvard go, and the majority of the rest end up at Yale, Princeton, MIT or Stanford.

The favorite talking point, which has a grain of truth to it, is that the quality of the lectures is not better than an average school. But it is the chance to be in a community where nearly everyone is amazingly smart (and the few who aren't are often the kids of the super-rich), and where you can do research under the top people in almost any field, and who have connections that get great jobs and/or grad school spots.

Harvard only costs $160,000 if you are from a rich family. With the $25 billion endowment and no money wasted on sports scholarships, the financial aid grants work out to free tuition for anyone from a family under $60,000, and more than 1/2 off tuition coming from a family with an income in the low six figures. The rest is covered by loans far below market rates.


This myth comes from the need of conservatives who need to hate all the colleges they were either rejected from, or were wise enough not to waste an application fee on in the first place. Time to let it go and move on!

rast said...

Is Geoffrey Miller your pen name, Steve? Great quote, great summary, but nothing new to anyone who's been reading your blog for more than 3 months.

AC said...

Striking. What would a competitive market in testing do in view of the two competing priorities?

Anonymous said...

Man, I can't wait to read this book. The author is on target, especially in terms of the restricted social contact high-IQ people have with each other. I administer IQ tests on a regular basis as part of my job; however, since I am administering them to federal pre-trial defendants as part of evaluations for competency and/or sanity, I get to see, day in and day out, how being on the left half of the bell curve can impact one's functioning in society. Believe me, I haven't run into too many defendants who would be sitting around discussing superstring theory, but for the lack of an "Idiot's Guide to Superstring Theory" that could teach to their specific learning skills.

Then, when I socialize with peers outside of work, these very statements (as identified by Mr. Miller) are made without a hint of self-awareness, by people who ought to know better. I feel like I'm in the episode of Seinfeld where the real world almost comes into contact with the Bizarro world; I want to take some of my high IQ friends and introduce them to my work clientele, for maybe a day or two, just to see how their world view changes should they deign to speak with someone 2-3 standard deviations below them on general intelligence (aside from, you know, getting their lawn mowed, having their oil changed, etc.).

Antioco Dascalon said...

ETS doesn't make LSAT (Law School Admission Council) or MCAT (Association of American Medical Colleges).

MQ said...

Ummm, the opening part of this basically acknowledges that G is a crock -- there is no corresponding single mental ability, it's an abstract social science modeling choice for a complex set of phenomena that is not well understood -- then the second half goes off into a bunch of elaborate conspiracy theories for why various people don't put much stock in it. How about: social sciences are notoriously imprecise and pseudoscientific, and social science modeling choices for stuff we don't understand too well are going to be heavily contested.

agnostic said...

"I have met theoretical physicists who claimed that any human could understand superstring theory and quantum mechanics if only he or she was given the right educational opportunities."

You see this a lot among Ashkenazi intellectuals --

"Well my blind uncle Murray can understand string theory! So anyone else should be able to, given the right education."

Most of their conception of a low-status man in the street is their older relatives who were low-status but intellectual.

Now, if they thought of the peasants who surrounded their ancestors back in the old country, I think they'd wake up and change their mind about everyone being a latent Cal Tech professor.

Anonymous said...

Bob said:

"Harvard only costs $160,000 if you are from a rich family. With the $25 billion endowment and no money wasted on sports scholarships, the financial aid grants work out to free tuition for anyone from a family under $60,000, and more than 1/2 off tuition coming from a family with an income in the low six figures. The rest is covered by loans far below market rates."

That's a load of BS re sports scholarships Bob. Ivies do give what amount to de facto sports scholarships by gaming financial aid. I know several (white, middle class, i.e., non-AA admittees) people who went to Ivy league schools on the basis of athletic prowess (football, basketball, swimming and raquet sports) who scored around 1100 to 1200 on the SAT (pre-recentering). These were not bad scores, but they were not in the range of the non-AA, non-athlete students. Some of these people came from truly wealthy families (far more than the Obama defined $250k/yr wealthy, which really isn't wealthy at all - see http://vdare.com/roberts/090329_obama.htm). They were able to negotiate back and fourth between several Ivies and other top schools who technically didn't give athletic scholarships to garner financial aid packages ranging from $16k to full cost (which back then was in the low $30ks). If they want you for their sports teams, not only will they drop their academic standards (a lot), but they will also give you more financial aid than they would a similarly situated regular admittee (including aid to people who normally wouldn't receive it).

Anonymous said...

This comment is directed at BOB,
I tried to do something similar with gre subject exams. They can't be used for purposes of employment. They will not be released to employers. I suspect the sat is similar. And I don't think anyone is saying to pass up a chance at attending harvard if you get the chance (Hell i'd go just for the scenery and amenities. Have you seen that campus?).

Anonymous said...

"I have met theoretical physicists who claimed that any human could understand superstring theory and quantum mechanics if only he or she was given the right educational opportunities."

My pa is such a physicist. His attitude towards me and my other siblings was that we were so obviously dumb compared to him that we should just take up normal jobs. Now, in midlife, we are battling our way through those "right educational opportunities" to correct the erroneous advice of this super-genius. Whatever they may say in public, and I know a few such people through church and family, most of them are very aware of their special IQ status and arrogant bastards on account of it. So their public announcements are not about envy-deflecting or idealism but rather to hide their obnoxious arrogance.

Truth(er) said...

"Anyone who could get into Harvard could also go to a state school or a solid private school for free, including living expenses, and then list their 1600 SAT score on their resume (and have a certified copy handy for employers). Yet something like 80% of people admitted to Harvard go, and the majority of the rest end up at Yale, Princeton, MIT or Stanford."

Because the person who has a 1600 on his SAT but who does not go to a top school will be put at a decided disadvantage compared to all of those who do. Employers will certainly regard the disjunction between the high SAT score and the lack of academic credentialism as a red flag. Try getting into Goldman Sachs without an Ivy background and see how far you get.

Besides, if employers would hire kids with 1600 SAT scores alone, then why would anyone bother going to college?

The university as a certifying agency is extremely important.

Bob said...

Responding to the comments with my name:

1. Sure official copies of standardized test score reports might not be released directly to employers (I wonder if this rule is enforced that well) but they will certainly send it directly to the person who took the test.

Some college transcripts also list SAT scores, in which case an official transcript mailed from the college will suffice.

2. Sucessful negotiation of financial aid packages at ivy league schools is not limited to athletes, but rather is very common.

Anonymous said...

OT: Steve, there is an interesting article at WND about the political donations of the infamous mortage pirates Herb & Marion Sandler. These are the a-holes who were parodied on Saturday Night Live with the caption "People Who Should Be Shot". But the kicker is the photo of the dynamic duo accompanying the article. In the posed studio photograph they look like either:

a) satanists
b) hired killers
c) morticians
d) suicide cult leaders

I mention this photo because it's a highly crafted and posed shot. Just bizarre.

Guess who's riding to rescue of those failing newspapers 'King and queen of toxic mortgages' underwrite 'independent' investigative reporting – with a tilt

Anonymous said...

Miller is taking a big risk with his career, writing these things. I wonder if he is angling for Pinker to recruit him to Harvard, so he (Pinker) won't be so lonely? They are both skating on the same thin ice.

Anonymous said...

agnostic,

"Now, if they thought of the peasants who surrounded their ancestors back in the old country, I think they'd wake up and change their mind about everyone being a latent Cal Tech professor."

They do think of the peasants who surrounded their ancestors back in the old country. Except when they do, everyone isn't a latent Cal Tech professor, everyone is a latent Nazi SS officer.

Anonymous said...

What I don't understand is that the vast majority of academics work in universities (like New Mexico) that might have incentives to go after the big brand names, because they're offering the same product at a lower price, so why haven't faculty at the hundreds of lower tier schools made Miller's point before?

Anonymous said...

"MQ said...

there is no corresponding single mental ability, it's an abstract social science modeling choice for a complex set of phenomena that is not well understood"

IQ derives almost all of its validity because it is a good proxy for psychometric g. Arthur Jensen reports (P. 91, The g Factor) a g loading of about 0.88 for most IQ tests. Jensen suggests that the word intelligence not be used in scientific discussions, because it lacks a scientific definition, and that we should instead focus on g, since it is unambiguously defined as the product of a hierarchical factor analysis. It happens that all categories of test items correlate with one another to at least some degree. The ultimate relationship between the various categories of cognitive activity is reflected as g, which is common to all mental abilities. Anyone interested in how g is defined and derived, should consult The g Factor, 1998.

Having narrowed intelligence down to whatever it is that is reflected in g, the next logical question is what is it that causes variations in g? Charles Spearman discovered g in the early 1900s and developed factor analysis, plus a number of statistical methods that are used in psychometrics and other fields. Curiously, Spearman was never satisfied that he understood the nature of g. Today, g remains somewhat difficult to comprehend, but there is now much evidence that it is based in physiology. Psychometric g is substantially heritable (rising from 70% in young adults, to over 80% in older adults). There are some identified genes; one example is Igf2r, which is associated with high intelligence. Although there are various physiological candidates that may account for some degree of what we see as g, two are especially salient: nerve conduction velocity (NCV) and myelination.

NCV correlates to g, with faster NCV in more intelligent brains. This is an area that has been extensively studied and reported by Arthur Jensen. While there may be various ways in which NCV affects mental activity, the one that Jensen most often mentions relates to the high volatility of information held in working memory. Faster movement of the information, implies that it can be used before it is lost.

Ed Miller noted that Jensen’s explanation of the role of NCV in intelligence does not explain the observation that the standard deviation of response times to external stimuli correlates negatively with IQ (smaller SD reflects higher IQ). Miller's paper, “Intelligence and Brain Myelination: A Hypothesis,” first appeared in Personality and Individual Differences, Vol. 17 (December 1994). By assembling a huge number of observations, Miller has been able to support his hypothesis that variations in intelligence is a function of the degree of myelination in the brain. His explanation for this is very lengthy, but is basically that myelin acts as an insulator preventing neural noise from disrupting brain signals in a manner that is somewhat like cross-talk that is seen in electrical circuits. As noise disrupts brain signals, the brain attempts to retransmit the information. If there is a significant loss of transmission integrity, there is a cascading of errors, overwhelming the brain so that it cannot progress. This is apparently what is seen in testing of people by increasingly difficult problems. A point is reached in which the person cannot solve the problem in any amount of time. They simply cannot manage the number of bits of information that must be manipulated for a solution.

Another strong indication that intelligence stems from physiology is that g can be measured passively with results that are as good as standard power IQ tests. This can be done by two quite different methods. One method is known as electroencephalography, or EEG. This method involves a variety of measures which are taken from averaging a large number of brain waves. The test subject need only be connected to the measurement equipment his brain waves are evoked by a clicking sound. Measurements are based on such things as the complexity (string length) of the average signals, and the zero-crossing points of specific parts of the typical response pattern. More intelligent brains produce a more complex form (greater string length). The technique is a completely passive observation of brain activity.

The second method, consists of a battery of very simple tests, known as Elementary Cognitive Tests (ECT). The test subject is asked to press a button when he observes a simple condition, such as the turning on of lights, or the playing of a sound, or the movement of a line on a projection screen. All such tests can be done by virtually all subjects in less than one second. Some of the tests involve comparisons or discriminations, but are not difficult to perform, even by very dull people. Sensors on the scalp measure the brain reaction time for each task. More intelligent subjects have shorter average brain reaction times (this does not include hand movement time). Likewise the standard deviations of higher IQ subjects is lower than for less intelligent subjects. Interestingly, the reaction time correlation is independent from the standard deviation correlation, strongly implying that they have different causes. Each ECT is somewhat g loaded, but when combined, the net measurement correlates to test results by standard IQ tests as closely as one IQ test correlates to another one.

The importance of EEG and ECT testing may extend into many aspects of understanding and measuring intelligence, but the most significant aspect of these modern laboratory techniques is that they seem to be more direct ways of getting to the physiological nature of intelligence. Both techniques seem to give support to Miller's myelination model, and both indicate that it is communication between widely separated regions of that brain that is central to the variation we observe in intelligence. As something of a side note, both EEG and ECT measurements show the same variations between racial groups as are seen with traditional IQ tests.

Among the other interesting aspects of physiological correlates to intelligence, more intelligent brains show the following:

· PET scans show that they are more efficient with respect to glucose uptake.

· MRI shows higher gray-white matter contrast.

· Brain volume is greater (correlation of +.44, after correcting for body size).

· In MRI examinations, there is a phenomenon known as T2 relaxation time, which is shorter. The importance of this is that the T2 relaxation time is an indication of the number of biological membranes in the immediate vicinity of the affected protons (in water).

a researcher said...

Ummm, the opening part of this basically acknowledges that G is a crock -- there is no corresponding single mental ability, it's an abstract social science modeling choice for a complex set of phenomena that is not well understood -- then the second half goes off into a bunch of elaborate conspiracy theories for why various people don't put much stock in it. How about: social sciences are notoriously imprecise and pseudoscientific, and social science modeling choices for stuff we don't understand too well are going to be heavily contested.you're wrong.

[note that my response has as much substance as yours.]

Anonymous said...

"....more accurate the tests are as indexes of general intelligence, the more biased they look across groups..."

That alone makes him a prime candidate for the mau-mau.

Anonymous said...

Off topic, but I can't help wondering how all this clamoring for a woman or minority on the SCOTUS will affect the Ricci case. If I were Souter I'd be insulted to put in 20 years on the bench only to find out activist groups saw me as clogging up a spot for someone with darker skin. If he takes his job seriously, he would want his replacement first and foremost to be competent, with the impact on diversity a secondary concern. Once that light bulb goes on, it's not hard to realize firefighters are in the same situation.

dearieme said...

"Whenever policy issues are important enough, we turn to the concept of general intelligence ...": so it's rather the opposite of the US Constitution, which is abandoned whenever policy issues are important enough.

Tony1 said...

Why is this not a breach of copyright? The book will not be released for sale until next week.

Anonymous said...

I think that the whole concept of "IQ" is of questionable worth. Far too many of those people with IQs over 140 are prone to amazingly stupid ideas which people with a far lower IQ laugh at. This is most notable in the fields of politics and sociology.

We'd all be better off if we accepted the idea of multiple types of intelligence. The type of intelligence associated with doing research into superstring theory is of very little utility in most spheres of life, and actually seems to be associated with a certain type of stupidity about many topics. Think of them as idiot-savants.

Anonymous said...

I think Miller has a point about bright folks thinking that those less intelligent just need better teaching.

Speaking for myself; I am quite bright so I suffer from the debility that Miller cites.

On the other hand, I know damn well that I am not as fast as some fast runners I have been around. They may think that if I practiced I might be competitive; but I know as a fact that that is not correct.

Those below, who cannot fathom how those above do what they do, understand in their gut and in their mind real diversity. Those above simply see that those below do not perform. That is a great difference.

I might say that when I went to college I got to know very well some folks who went on to academic careers in abstract mathematics. I had no idea what they were doing or how they did it; even though I majored in physics and applied math and went on to a successful career. Same point.

Pat Shuff said...

Alimentary my dear Watsoned, alas poor Geoffrey we barely knew ye. Oh well, to the rack with you.

Horrors, brain size?

Miller, G. F., & Penke, L. (2007). The evolution of human intelligence and the coefficient of additive genetic variance in human brain size. Intelligence, 35(2), 97-114. [PDF]


http://tinyurl.com/ccpgb3

Benn Gunn said...

Hmm, most of the readers interested in intelligence are anonymous. I propose the intelligent are a snobbocracy, secure to the point of impunity, dealing with the less clever rather cruelly and with malice. The evidence? the typical stultifying textbook.

Mr. Anon said...

"Bob said...

The idea that very selective schools are just legal ways to give potential employees IQ tests is such BS.

Anyone who could get into Harvard could also go to a state school or a solid private school for free, including living expenses, and then list their 1600 SAT score on their resume (and have a certified copy handy for employers)."

And how many people list thier SAT or GRE scores on thier resume? Do you?

Anonymous said...

Anon, thanks for the information on myelination and EEG tests of g.

Steve could you put together a recommended reading list on the subject of intelligence? Normally I just troll on this blog, but despite my best efforts I have learned quite a bit. A short list of around 5 books would be helpful. A post on the subject could separate the books by difficulty, or you could have two or three sections, one section for each subject you consider important.

Jim O said...

I wish I could find a copy of a documentary from CBS Reports aired in 1975, hosted by Dan Rather called "The IQ Myth."
I haven't seen it since it was broadcast, but I remember it as probably the slickest propaganda piece I've ever seen. Today, CBS never could have gotten away with this laughably lousy argument that IQ does not exist. If only there had been such a thing as blogging back then.

Anonymous said...

Bob said:

This myth comes from the need of conservatives who need to hate all the colleges they were either rejected from, or were wise enough not to waste an application fee on in the first place. Time to let it go and move on!
Oh wow. Projection.

BigWaveDave said...

It is a truism that many cloistered academics "believe" that anyone can learn complex things, but theirs is a belief of the comforting variety, one that blends in well with the other decorative delusions of their environment. A quantum physicist might "believe" that Bell's Theorem is within Joe the Mechanic's grasp, but he "knows" that swapping out a transmission is not something he himself can be taught. At all levels of our society the recognition of innate inability is enthusiastically accepted provided the inability is for a task considered the realm of a lower social rung.

clem said...

Why is this not a breach of copyright? The book will not be released for sale until next week.Uh, maybe because copyright violation has absolutely nothing to do with whether or not a book is in print. Copyright is established in North America at the time of creation of a work; whether or not it ever gets into print is irrelevant to the principle of copyright infringement.

The most Steve's quoting could be would be a violation of "fair use," if it was judged to be too long of an extract. More likely, though, the author and publisher would recognize that Steve is giving them some damn good PR and buzz here, to a very targeted audience.

Plus, since the book isn't for sale yet, how do you think Steve got ahold of a copy of it? 'Cause right now, only the author, the publisher, and reviewers they've specifically chosen to send copies to, should have access to it, eh? That is, Sailer couldn't have just walked into a bookstore and bought it off the shelves, any more than you or I could.

So where do you think he got his advance copy from?

Anonymous said...

re: posting your SATs, etc. on a transcript

The NALP (Nat'l Assoc. of Legal Professionals) bars law students from putting their LSAT scores on their resumes and cover letters.

-bushrod

Mr Lomez said...

A quick word on the "New SAT":

In 2004 the College Board introduced a new 3rd section to the SAT, consisting of a 25-minute essay, and 2 multiple-choice sections that (ostensibly) test students' understanding of basic grammar rules.

Because of this additional 800 point Writing section, I have to reluctantly concede that one's score on the New SAT -- just as the folks at Harvard would want you to believe -- is now marginally, if at all, a reliable proxy for IQ (or g, or whatever you feel most comfortable calling it).

I am owner and head-instructor for an SAT company in a major metropolitan area, and I see this unsettling reality play-out on a weekly basis. Specifically, my Chinese, Korean, and Japanese students -- largely the sons and daughters of first generation immigrants (I am in a city, not a suburb) -- with otherwise relatively high IQ's, lack the intuitive understanding of Standard English that other students with less high IQ's possess. Even when their vocabularies are commensurate with their relative IQ, these immigrant/ESL students have trouble grasping certain concepts particular to English's quirky -- and often arbitrary, bizarre, and counterintuitive -- constructions.

Two examples of grammar rules that confuse the hell out of these kids:

1) Prepositional idioms. For instance, that one speaks "to" another, rather than speaks "with" another,
2) Either/or sentences, in that the verb pursuant to the main clause only pertains to the last in the list of antecedent subjects.

Kids who grew-up in Standard English speaking households, know this stuff without knowing it -- even the dumb ones. The much brighter, more college ready, more sophisticated thinking immigrant students, simply don't (at least, not until they take my class :).

The essay is no better. Though we can all agree that the ability to articulate a meaningful and cohesive argument is essential to college success (and correlative to IQ), the essay on the SAT is graded according to a vague, and therefore useless, rubric that boils down to, "does this *sound* good." There is no credence given to the complexity of one's argument, nor to its depth of subject-matter (an essay about the Visigoths is no more score-worthy than one about American Idol contestants). Again, the people who suffer from the essay are high IQ students who happened not to be rhetorically gifted, either because English is foreign to them, or because they haven't been schooled in the art of bullshit.

The upshot of all of this is that the College Board, in conjunction with ETS (though the exact relationship is unclear), has effectively, with this new 800 point section, eliminated the 1 to 1 correlation between SAT scores and IQ -- if you believed that there ever was such a correlation.

The interesting question is, why did the Powers That Be deem it necessary to distort the meaning of the SAT? And why, in all the ways they could've distorted it, did they choose this way in particular?

[half-baked theory] I can only guess that this new section of the SAT was a direct assault against the influx of east Asian (and possibly Indian) students at the country's most selective universities. It is no secret that Asian students are extraordinarily high academic achievers. And because of their huge numbers -- ever increasing due to China's booming (or once booming) economy -- the amount of qualified east Asian students for the limited number of Ivy League spots, must have been reaching a point of critical mass.

How do you stem that tide? What we also know about these high-achieving east Asian students is that their English skills are less than substantial. These English skills, rather fortuitously for the P.T.B., are hard to "learn," regardless of IQ and especially for east Asians, because of the profound differences between English and their L1 tongue. Therefore, a 3rd section of the SAT that tests exclusively for those skills, will surely mitigate whatever advantage those students had on the original 2-part test.

And so it shall be. [/half-baked theory]

Soul Searcher said...

Anon of Myelination,

Thank you very much for that. g seems, from what I can gather, seems to be be closely analagous to processing speed or the idea of say, "clock frequency" in personal computers. But even this "fundamental" quality of g shouldn't completely specify the underlying architecture of emergent intelligence, correct? I mean, it should be highly unlikely that the modular components of intelligence evolved similarly in different groups, which could allow individuals of similar IQs to have quite different distributions of problem-solving strengths, right? Hearkening back to computer analogy, several of the improvements Intel made in the Willamette CPU involved changes in the micro architecture of the P6, rather than finding ways to up the clock speed, which lead to measured increased performance. If I'm even making sense here, is there any research on that type of idea?

ben tillman said...

The NALP (Nat'l Assoc. of Legal Professionals) bars law students from putting their LSAT scores on their resumes and cover letters.Hardly. The NALP has no authority to bar anything. However, putting your LSAT on a resume is generally an unwise decision, though I vaguely recall seeing it done once or twice.

Look, Smithers... said...

I, too, breathlessly await this book. It will change everything. The whole structure of lies will come crumbling down. My people will be redeemed. We will no longer suffer dispossession or mockery. Our history will be written by us. We will no longer be slaves in our own lands. This is finally it. I can't believe I'm alive to see it. The book is coming! A mildly conservative book is coming!

Svigor said...

Think of them as idiot-savants.Middle classspeak for "lotsa book learnin', but he cain't drive a tractor for shit!"

(Actually, the Standard Redneck English for this is "book smart")

I mean for Chrissakes, really. Go two standard deviations or so up or down and there's no relating, except in a limited way. People with 90 IQs really don't know WTF I'm talking about unless I talk down to them.

Svigor said...

Bob Hume has a point that people "g-pomorphize" those around them into being up to tasks they are themselves up to. Interesting post.

But let's face it, anti-realism is popular because people are trained to believe it. (smarter dogs are generally easier to train, right?)

Anonymous said...

"Hmm, most of the readers interested in intelligence are anonymous."


That's to avoid the $PLC lawsuits

Anonymous said...

Mr Lomez, I have a question: Are the new SAT math and critical reading sections all that different from the verbal and math sections of the old SAT? Didn't they really just append the SAT II writing test to the SAT? Before the changes, applicants to elite colleges generally had to take the SAT and the SAT II writing. Will appending the latter to the former really change the way colleges evaluate applicants? I can, however, buy your argument that the changes in the SAT may be an attempted "solution" to the SWPLs Asian competition "problem" as plausible if they do indeed evalute the test scores differently now that they are combined.

Charles Murray was of the opinion that you don't really need the SAT because SAT II achievement tests are just as heavily g-loaded and carry less cultural baggage from leftist disinformation with them. However, by adding the writing component to the SAT they are effectively doubling the verbal component (where Asians are retively weak) compared to the quantitative/visuospatial component (where Asians are relatively strong).

Anonymous said...

Mr. Lomez makes great points. I got a 1560 on the math/verbal of my GRE last year, but only a 3.5 on the essay. I think it had something to do with the fact that I decided to write about the evolutionary differences between populations in shaping culture.

Acilius said...

"Of course, such scientists talk only with other physicists with IQs above 140, and seem to forget that their janitors, barbers, and car mechanics are in fact real humans too, so they can rest comfortably in the envy-deflecting delusion that there are no significant differences in general intelligence."

I think it's more likely that they know very well that saying "Superstring theory is so easy anyone should be able to understand it!" signals that "Superstring theory was so easy for me that I can't imagine anyone having difficulty understanding it!" Which if true would be a lot more impressive than an admission that the speaker has reached his or her understanding only as the result of long hours of hard study.

Mr Lomez said...

Anon,

Unfortunately, the new Verbal section has also changed for the worse. For one, there are no more analogies. I am not a psychometrician, but I would hazard a guess that the analogies were perhaps the most "g-loaded" of any of the original test's Verbal question types. To make matters worse, rather than replace the analogies with additional vocab questions, they've instead included additional "passage-based reading" questions. The passages and their corresponding questions are so fraught with internal contradictions, obfuscated language, and PCness, that it's border-line laughable. I can assure you that these questions do not test how well a student can comprehend a given text, but rather how well they understand the conventions of the SAT test itself. To wit, I can teach my students how to automatically disqualify at least 2 answer choices per question, based exclusively on the presence of certain "red-flag" words therein.

The Math section, apart from some trivial differences, is the same.

As for the SAT II Writing requirement for Ivies...I'm fairly certain that you only need to take 2 or 3 SAT Subject Tests, but not necessarily Writing. Hard-science applicants (i.e. - the east Asian applicants in question) probably would be expected to submit SAT II scores only in Math, and then some combination of physics, bio, chem, (and maybe Studies in Rice-Paddy Ancestry, if and when such a test is offered) ---- Although, I'd argue that this point is kind of moot. Most of these kids that we're talking about, unsurprisingly, work their tails off to get through AP English, despite their language setbacks. So college admissions boards need only look so far to see that they satisfy -- at least -- the minimum level of English/writing proficiency required to succeed as a university scientist (which, based on what we know about academic science writing, is not a particularly high standard anyway).

I'll add that I totally agree with Murray. The SAT's are kind of stupid, wholly irrelevant in light of the SAT II's, especially given the 2004 changes. That said, it won't stop me from making money from them. Make no mistake, they can be taught.

Anonymous said...

"(Actually, the Standard Redneck English for this is "book smart")"

Whatever you chose to call it, the fact remans that lots of these "very smart people" are very smart only in extremely narrow ways and that outside of their area of expertise, string theory for instance, they are considerably less smart than the guy fixing their car or their plumber.

Auto mechanics and plumbers are smart enough not to fall for communism or fascism. If only the same could be said of people with Phd's in physics.

SFG said...

"Auto mechanics and plumbers are smart enough not to fall for communism or fascism. If only the same could be said of people with Phd's in physics."

The mechanics and plumbers fell for that stuff in Russia and Germany. And Germans, by most measures, seem pretty brainy. Most likely they were desperate in the Depression and bitter over WWI.

The 'intellectuals lean to the left' thing probably has more to do with peer groups than anything else.

Anonymous said...

most of the readers interested in intelligence are anonymous. - Benn Gunn

That your real name Mr Gunn?

We are talking heresy here, no-one wants to get burnt at the stake do they. Not a risk orthodox liberals have to run, though they do enjoy seeing themselves as brave transgressives speaking truth to power etc etc

Ray Midge said...

And how many people list thier SAT or GRE scores on thier resume? Do you?,This just isn't done.

I went to a very well regarded law school. When you interview for summer positions for your 1L summer break, you do it early in your 1L year. You don't have grades yet by which you can distinguish yourself from your classmates.

Using whatever I could, I listed my high-ish LSAT score, thinking that'd help me stand out.

After one not-especially-good, not-especially-bad interview, the guy takes a moment to tell me on my way out that I really needed to take that off. Just not done. Might've en used the work "tacky." Don't remember.

So there you go. It's "just not done." I'll leave it to you to discern why the culture would evolve a "just not done" mentality with regard to publishing one's LSATs on one's resume.

Anonymous said...

Perhaps the problem we face is that its not mechanics or plumbers - or physicists for that matter - who come to dominate politics and public policy. Its graduates of law, politics, english and economics etc

rainy_day said...

Re: physicists thinking everyone could understand string theory - it's just egocentrism (assuming the world looks the same to everyone else as it does to you), everyone does it until forced to do otherwise (clear in young children). Eg I heard on a show some special forces say, "It's not that we're so great, just that everyone else sucks". It's natural, you model everyone's level of skill as that of your own, and assume everyone else just doesn't have the same drive or whatever. Now for us dummies who've been confronted more often with the fact that our brains generate more heat than light, we can 'imagine down' more readily.

Anonymous said...

"The 'intellectuals lean to the left' thing probably has more to do with peer groups than anything else."


I think it has to do with the fact that "intellectuals" tend to believe that all the worlds problems can be solved if only smart people (namely, themselves) are given complete control.

The complete balls-up they consistently make of things never seems to dent their self-confidence though.

Conservatism, once upon a time, was opposed to the whole idea of a society run by a technocratic elite. So the left was the natural home of the intellectuals.

Anonymous said...

"I think it had something to do with the fact that I decided to write about the evolutionary differences between populations in shaping culture."

Brave or stupid?

MQ said...

[note that my response has as much substance as yours.] No, your response just shows you don't get it and feel like being snarky about it. IQ = G, and G is just a referent for a bunch of correlations we don't understand. Try here for further explanation.

Take any bunch of measures of anything, and you can extract a common factor. That's all G is. THe kind of out-of-sample correlations that Anonymous pointed to in the post before yours do start to get you somewhere. But only if they are part of a real theory of how intelligence works in the brain, which we are a very long way from having.

Once we truly understand intelligence -- instead of having a bunch of Mensa types trying to feel superior on the web -- I suspect we'll find lots of contradictions and tradeoffs within it, with many mental abilities negatively correlated with other mental abilities and forms of intelligence. (And others being positively correlated, of course). Common sense observation certainly supports this.

Svigor said...

"(Actually, the Standard Redneck English for this is "book smart")"

Whatever you chose to call it, the fact remans that lots of these "very smart people" are very smart only in extremely narrow ways and that outside of their area of expertise, string theory for instance, they are considerably less smart than the guy fixing their car or their plumber.

Auto mechanics and plumbers are smart enough not to fall for communism or fascism. If only the same could be said of people with Phd's in physics.


Speak for yourself. I've rubbed elbows and worked with with many rednecks, and I'm confident I could best the vast majority at any measure of general intelligence. And I say that as an honorary redneck.

Svigor said...

But I will admit that the human mind (or at least, mine) isn't well-suited to multi-tasking. When I do some relatively boring, menial task, my mind wanders, and that can introduce dumb mistakes that typical rednecks wouldn't make. I suspect that this is because the task is enough to keep them interested.

Shah Doobie said...

I am owner and head-instructor for an SAT company in a major metropolitan area, and I see this unsettling reality play-out on a weekly basis. Specifically, my Chinese, Korean, and Japanese students -- largely the sons and daughters of first generation immigrants (I am in a city, not a suburb) -- with otherwise relatively high IQ's, lack the intuitive understanding of Standard English that other students with less high IQ's possess.

"Unsettling reality?"

Hey, Mr Lomez, the solution to your angst is simple: Move yourself to your preferred Asian country where all the gifted children are truly top quality students instead of undeserving second tier white bread. Of course, you'll likely find yourself in a monoracial society bordering on racial facism (China, Korea, Japan etc), but your mind will be at ease, instead of "unsettled", because the truly deserving students will not be held back by any language barriers.

What an educator you are. I'm sure, if you only could, that you'd have the entire world ruled by a "meritocratic" Asian elite in a heartbeat. Yes, from your experience, it is obvious that Western Civilization was created and sustained for millenia by second rate minds: a bunch of non-dynamic, non-creative, small minded, thieving hillbillies apparently stole every good idea they ever stumbled upon.

Thank Darwin there are no stubborn, backward, thieving hillbillies in Asia! Let's all thank Darwin that Asia has lit the lamp of Enlightenment, and shown humanity The Way and The Path through the darkness for all of these centuries.

Yes, it sounds like we can assume that you'd have non-Asians replaced by Asian immigrants in every nation and then the entire world could finally reach its true maximum intellectual potential: to be Asian in thought and action!

It sure is good to know that you're only one soldier in an army of educators spread out across North America, Europe and Australia, all diligently working to conquer the obstacles and usher in the new, superior Asian populace that will finally pull the white Western world out of its permanent dark ages.

Anonymous said...

All this bunk about how elite schools give you a chance to socialize with smart peers, etc., is precisely why employers won't use SAT scores instead of university degrees.

Using SAT scores would be a problem for the exact same reason that using IQ tests would be a problem--disparate impact leading to legal liability.

But since people like you make a fetish of credentials, employers can use them without admitting that they're actually just testing IQ in a roundabout way.

Mr Lomez said...

@ Shah Doobie

Did you sleep alright last night?

I think you're rearranging my argument to vent some other, unrelated aggression.

My point was that the SAT, in its latest form, is not a reliable proxy for IQ -- and this point is most obvious when you consider the relatively low performing, but otherwise high IQ east Asian students.

You write: "I'm sure, if you only could, that you'd have the entire world ruled by a "meritocratic" Asian elite in a heartbeat."I'd amend that statement thusly: "I'm sure, if you only could, that you'd have the entire world ruled by a "meritocratic" _____ elite in a heartbeat."I really don't care who's going to Harvard and Yale, so long as they're qualified. Specifically, I think, especially in the hard-sciences, our best schools should be recruiting students with the highest IQ's. Period.

I'll add that the New SAT hurts not just east Asians, but also speakers of Black English, non-white hispanics, speakers of White Southern dialects, et al. The only people this helps are white middle and upper-class students who come from already highly educated families.

Again, I don't care who's getting into the Ivies, or into the highly impacted and highly selective UC's, so long as those students are deserving.

Martin Walker said...

This is an interesting perspective. One-sided, but interesting. It calls into question the value of higher education, and it is good for us to revisit that question on a regular basis.

I would think that the recent finding that intelligence isn't fixed might help IQ become a less touchy subject. Just as weight-training transformed the weakling into a muscle-man in the general media thirty years ago, now brain training can increase a person's IQ.

Susanne Jaeggi and Martin Buschkuehl's study on Improving Fluid Intelligence by Training Working Memory (PNAS April 2008) recorded increases in mental agility (fluid intelligence) of more than 40% after 19 days of focused brain training.

I was so impressed that I contacted the research team and developed a software program using the same method so that anyone can achieve these improvements at home.
Mind Sparke Brain Fitness ProMartin
www.mindsparke.com
Effective, Affordable Brain Fitness Software

David said...

Martin Walker, that's nothing. Alyssa Rosenbaum ("Ayn Rand") claimed a person's IQ could be increased from 110 to 150 through "rational thought" and willpower. (In the book "Ayn Rand Answers.")

Cost of paperback "Atlas Shrugged" = about $8. Cost of your system = $46.95. You fail.

Anonymous said...

Who let the spam through?

Half Simga said...

"George Will has written two recent op-ed pieces addressing the legal issues that explain why cognitive tests aren't used by employers -- you guessed it, lawsuits due to group differences in outcome."

NO, the real reason is that HR people believe the liberal politically correct viewpoint that IQ doesn't mean anything.

The really do seem to believe that graduating from college CAUSES people to be smarter.

No mas Lo mez said...

"I really don't care who's going to Harvard and Yale, so long as they're qualified. Specifically, I think, especially in the hard-sciences, our best schools should be recruiting students with the highest IQ's. Period."

"Again, I don't care who's getting into the Ivies, or into the highly impacted and highly selective UC's, so long as those students are deserving."

Qualified. Deserving. These adjectives change depending on the culture, the era, and the race. What is "qualified" and "deserving" to you means nothing to in a Muslim land, nothing in 1920s Russia, nothing to postmodern America and the Hindoo, Israeli, Chinese, or Armenian inhabitant when he's looking to hire in his firm.

Man, it's like the slightest arc of the education industry touches someone and they're blinded by the ethnocentric reality that moves most of the world.

Anonymous said...

Martin Walker - thanks for the link there.

Well, looks like the whole bell curve thing is wrapped up there then. With a year or two the whole NCLB and IQ & the wealth of nations issues will be history.

What will we talk about then?

freak show said...

To mq:

cozma has been critiqued pretty extensively after publishing his incredibly complex work on 'g.'

here is a good takedown:

http://idiolect.org.uk/notes/?p=653

from the critique:

'The main thrusts of their arguments are that test data do not statistically support a g-factor. Gould’s argument is statistically incompetent (for a statistican’s critique see Measuring intelligence: facts and fallacies by David J. Bartholomew, 2004). Shalizi’s criticism is incredibly sophisticated, but likewise incorrect. In a nutshell, Shalizi is trying to argue around the positive correlations between test batteries. If those correlations didn’t exist, his argument would be meaningful. However, as I noted above, these intercorrelations are one of the best documented patterns in the social sciences.'

Anonymous said...

>"Although nominally dedicated to the highest standards of test validity, ETS is also under intense legal pressure to create tests that “are free of racial, ethnic, gender, socioeconomic, and other forms of bias.” This means, in practice, that ETS must attempt the impossible"

What you're referring to is checked by an analysis called differential item functioning. The difficulty of questions on tests are estimated, and odds of a given test taker getting each question right given their level of theta (ability) on the overall test are calculated. For example, a person with a high level of ability who did very well on the test would have 90% odds of getting an easy question right, a person of moderate ability would have a 50% chance of getting a moderately difficult question right, etc.

Test takers are then graphed by race, gender, etc, to see if a given group of people has erratic performance on a given item. For example, a question involving listening performance could exhibit strange patterns when used on Japanese test-takers, because a key word in the passage uses an "l" that could be confused by that sub-group as an "r", so that they respond to it in patterns we would not predict given their ability. Or indian kids could do surprisingly worse on it because it assumes a knowledge of American history that they do not have.

However, it should be pointed out that group differences in ability do not matter one whit to this process; it is inconsequential if blacks or any other group as a whole perform poorly on the question provided that their level ability is correspondingly low. It is only important if their responses differ from what we would expect to see. Indeed, it is just as problematic if a sub-group scores higher on a difficult question, as it could mean that answering the question right then comes to correlate negatively with success on the overall test.

Indeed, one would think proponents of "g" would applaud such a process. Because if "g" truly is an identifiable, unidimensional construct, the items that tap it well should exhibit invariance, and predict its presence across race and gender, even if some races had lower levels of it, on average.

It's disturbing that you do not seem to know about this. For lack of any credible academics willing to continue to pursue scientific racism, you and your little blog have been elevated to forefront status in dwindling community. Yet you seem ignorant of the statistical analyses that you assume would hold up your theory. You claim psychometrics holds the key to human worth- how about you educate yourself to the basics of it before you run your mouth again?

Sideways said...

It's disturbing that you do not seem to know about this. For lack of any credible academics willing to continue to pursue scientific racism, you and your little blog have been elevated to forefront status in dwindling community. Yet you seem ignorant of the statistical analyses that you assume would hold up your theory. You claim psychometrics holds the key to human worth- how about you educate yourself to the basics of it before you run your mouth again?

you've A) confused Sailer with the person he is quoting, B) confused different types of attempted norming of the test, C) made a simply and completely false statement in your last sentence. Good job.

Anonymous said...

Well, If you have ever considered writing some script to automate those things that you are doing many times on a daily basis, then you have a great and easy way to automate the entir e process. It comes without saying that a few imacros and firefox add-ons will automate all actions connected to a browser (Mozilla). If you wish to automate your system as such, then you may need another script, easy commands can automate your system. Even a beginner can automate things using a simple software called as sikuli all you need to do is to tell Sikuli what to do by giving some screenshots and easy commands.
A GUI can be used with Sikuli. Sikuli is an open siurce scripting application that will use a mix of simple commands like click, type, wait and so on. There's no internal API support, it just searches the screen for the image in the screenshot-that implies that you'll use it for anything. There is literally no limit on how you utilize it to automate things.
This is too good to be true, but if you watch the following video, you'll come to know that this is very easy. If you visit their home page, you get a lot of tutorials and support; it will be a cake walk even for beginners. You'll be able to check the video as well as get the software at [url=http://technoages.com/operating-system/apple-mac-os/automate-everything-through-a-simple-script-make-your-computer-listen-to-you/]TechnoAges.com[/url]