Talk:Race and intelligence/Archive 42

This is an archive of past discussions. Do not edit the contents of this page. If you wish to start a new discussion or revive an old one, please do so on the current talk page.

Archive 35

←

Archive 40

The purpose of this article

On the first note, sorry for deleting the section beneath this, but I really dont think it has any place in this discussion page.

Anyway, I really think this article is confusing as to what its purposes are. Is it to inform about the actual differences, statistic or scientific (no matter how good data and material we have concerning this issue in those two categories) about race and intelligence, or is it to explain why there are such differences and discuss if there even is? The article seems to be more of an attempt to explain and discuss why instead of showing the hard facts, that is, what we can nominate as the closest to hard facts, the best data there is.

It might just be me, but I had hoped for a topscore list of statistically most intelligent races, and thereafter a brief argumentation of the statistical evidence, the terminology and definitions, that is, the measurement tool of intelligence, and the definition of race.

An attempt of explaining or discussing why should be the uttermost external subject of the article, the last section. Why? Because thats the most debateable subject, and probably the most controversial, even though the others are pretty controversial too.

But this article is more of a mess, a mix of explaining, discussing and argumentating for why this is, while brief details about the actual (most likely) statistical and scientific facts are revealed bit for bit whilst the mess of explanations goes on.

Is it just me? Or is there an excessive prioritizing and amount of space used on discussing explanations in this article, instead of focusing more on the actual facts, that is, the most likely facts?

- EMB —Preceding unsigned comment added by 85.82.195.131 (talk • contribs)

Explanations are the product/aim of science, and to the extent that this article reflects what scientists have said about this subject, this article too will focus on explanations. If you just want data tables, see Race Differences in Intelligence and similar books. --W. D. Hamilton 00:55, 22 December 2006 (UTC)

Data tables are often the most controversial part of this subject, with strong disagreement over the consistency of definitions (one drop rule, for example), and even the measurement of "g". If we were to simplify this article to "a topscore list of statistically most intelligent races", we could come up with half a dozen different lists, each based on a different definition of intelligence or race. I might have a list with only one race, called "human", whereas someone else may have a "race" for every 200 square mile patch of dirt on the planet.

I suppose the "most likely facts" are that there is no significant gene-mediated differences in intelligence between arbitrary categories of race, and that the very concept of "intelligence" is complex enough not to easily lend itself to simple data tables (for example, explain to me what you ate for lunch yesterday using only two letters of the alphabet - trying to describe intelligence on a single scale is probably just as accurate). Large scale intelligence studies with full DNA patterns of each participant have not been conducted, and AFAIK, none are planned. That being said, since there is great disagreement as to whether that is "most likely" on a number of fronts, a deeper discussion is required.

I'm sorry this doesn't match your expectations, EMB, but I'm afraid that you've come across a topic that does not lend itself to easy answers. --JereKrischel 07:57, 22 December 2006 (UTC)

This article is based on a pseudoscience; please be aware of this before adding to the discussion.

This article was nominated for deletion for obvious reasons. This article has questionable scientific value, which is why it was nominated for deletion. Please view the discussion points above to understand why the article was nominated for deletion. Making a judgement before understanding how the brain works makes this article an assumption, or a non scientific way of looking at things. Until scientists can actually understand how the brain works this article will only be based on an opinion that cannot be verified by modern technology or devices. One circumstance has shown that people with half a brain can be successful academically, see Ahad Israfil . Which proves that not only a smaller brain can perform well, but also brains that been shot in half have been known to pass college and graduate with honors. Jon Dec. 18 2006 —The preceding unsigned comment was added by Jon568 (talk • contribs) 23:06, 18 December 2006 (UTC).

Thank you for your opinion. Of course, we should also remember that this may not be the opinion of the many scientists who have contributed valuable research to this subject. – Agendum 23:35, 18 December 2006 (UTC)

It's mainstream science, the APA has said so, the biggest names in the field say so. Perhaps you're starting from a dogmatic certainty that it can't be science so it isn't? The data certainly disagrees with you and so do essentially all specialists in the field. —The preceding unsigned comment was added by 70.91.235.10 (talk) 22:48, 26 January 2007 (UTC).

Actually, it depends on which researchers in which field you mean. "Race and intelligence" means psychologists and psychometricians, but also biologists, anthropologists and several other specialties which may not agree to what you find "mainstream".--Ramdrake 23:04, 26 January 2007 (UTC)

That's true; scientists can believe whatever they want to believe. When scientists figure out how the brain actually works and can differentiate brains by race scientists will then base their data on scientific facts, not pseudoscience which is represented in this article. Jon Dec. 18 2006

20 to 80 percent

the notion that substantial heritabilty = 20 to 80% crops up all over the place. here's a passage from a paper i read recently:

Using different methodologies, behavioral geneticists have consistently found that [mental traits] heritability estimates range from about .2 to about .8, meaning that 20% to 80% of the differences between individuals in [mental traits] are accounted for by differences in alleles between people.

so i think there's a solid body of literature on which to rest the claim that heritable/partly-genetic = 20-80%, all environment = 0-20%, and all genetic = 80-100%. that is, i think we can feel confident describing the debate in these terms. --W. D. Hamilton 00:47, 16 December 2006 (UTC)

I think that's exactly what we're doing by being specific about the numbers...but I would also refer to you to the disagreement as to what "heritability" really means, in regards to the 20-80% figure - that is to say, "accounted for" meaning "correlates to" is very different than "accounted for" meaning "is directly caused by". In any case, I would hesitate to label those categories the way you've described them, and instead say "20-80%, primarily genetic", "0-20%, negligible genetic" and "80-100% overwhelmingly genetic". It seems to unfairly prejudice the "0-20%" and "80-100%" crowd by labeling them in terms of extremes, when in fact there is room on both positions to accept a small contribution by the other. --JereKrischel 09:24, 16 December 2006 (UTC)

If we are going to quantify these levels of contributions as percentages, they should definitely be cited, including citations that propose other percentages. I also agree that these labels seem misleading and unequally applied. Very few people are arguing 100% either way, and the article suggests "partly genetic" for the hereditarian position, but "environment only" for the opposing position. Whatever term is used for one should be used for the other. Jokestress 08:14, 22 December 2006 (UTC)

I'm glad this was asked because it is a mistake that perhaps others are making. Look again at the results from the Snyderman and Rothman survey.

This is perhaps the central question in the IQ controversy. Respondents were asked to express their opinion of the role of genetic differences in the black-white IQ differential. Forty-five percent believe the difference to be a product of both genetic and environmental variation, compared to only 15% who feel the difference is entirely due to environmental variation. Twenty-four percent of experts do not believe there are sufficient data to support any reasonable opinion, and 14% did not respond to the question. Eight experts (1%) indicate a belief in an entirely genetic determination.[110]

Any reading of Jensen, Murray, Herrnstein, or the other hereditarians will reveal that they believe there is "some" genetic contribution and "some" environmental contribution. The 20% threshold was identified by Reynolds (2000) as a sensible interpretation of the difference between "some" and "none", and the value of 20% is one that many authors converge on in the wider setting of the ability to detect "any" effect in behavior genetic studies.

On the other hand, I see no reason to qualify every instance of "partly genetic" and "environment only" with percentages. It can merely be pointed out when those terms are defined. --W. D. Hamilton 17:52, 22 December 2006 (UTC)

I have stated previously that I feel the Snyderman and Rothman is given undue weight in the article, but besides that, the problem still stands in application of labels. They report (in order of prevalence):

combination of heredity and environment (53%)
insufficient data (28%)
environment only (17%)

("Heredity only" is a very small number as I recall). "Partly genetic" could as easily be called "partly environmental," which I think gets at the issue of presentation bias I am concerned about. Jokestress 18:42, 22 December 2006 (UTC)

This issue has more than sufficient context given what's been published to show that the main question is whether there is some or no genetic contribution, with those arguing for "some" always making a positive case for the need for genetics as an explanation. To say "partly environment" is to obscure what the other part is, whereas "partly genetic" is quite clear. "Insufficient data" isn't an explanation, it's a non-explanation -- an inability to choose between alternatives, not a third choice -- it's like an "undecided" voter. If asked whether the gravitational constant G = 6.67E-11, you can say "yes", "no", or "I don't know"; but only "yes" and "no" are statements that have any relevance to what G really equals, whereas "I don't know" is merely a statement about the state of one's knowledge. --W. D. Hamilton 23:05, 22 December 2006 (UTC)

Am I wrong, or is there something screwy about Wdhamilton inferring from the 20-80% heritable claim that therefore "all genetic" is 80%-100%? Isn't this mistaking the y axis for the x axis? Slrubenstein | Talk 13:18, 22 December 2006 (UTC)

I don't get it. --W. D. Hamilton 17:52, 22 December 2006 (UTC)

What do you mean by "all environment = 0-20%, and all genetic = 80-100%?" I don't get that. Slrubenstein | Talk 14:52, 23 December 2006 (UTC)

Debates over removing things

The gap in black and white test scores in America remained stable

Claim: "However, research has shown that the gap in black and white test scores in America remained stable throughout the 20th century, fluctuating by less than one point"
Source: http://newswire.ascribe.org/cgi-bin/behold.pl?ascribeid=20061109.113857&time=13%2046%20PST&year=2006&public=0

The article that's linked to says: "Shuey’s (1966) review of the literature, which shows that Black-White IQ differences in the United States have remained at 15 to 18 points, or 1.1 standard deviations, for nearly a century"

Reason why:

This quote doesn't prove the article's claim unless Shuey's disputed findings are proven completely valid.
Shuey's study was done in 1966, so it's not exactly representative of a century, and doesn't consider new data, criticisms and methods since then.
The article criticizes Flynn's data (which doesn't show a 1 point change), but other scientists would dispute its criticisms.

--Urthogie 17:36, 23 December 2006 (UTC)

The quote supports the articles claim
It appears the article is referring to tests going back to the 1880's
If other scientists dispute the claims, add a sentence that shows that fact

Mytwocents 04:16, 24 December 2006 (UTC)

The claim doesn't even mention Shuey-- it just accepts his scientific findings as fact.
The claim about the "fluctuating by less than one point" can be found in the article to be based on Shuey's 1966 study. Prove me wrong if you disagree.
The problem is it doesn't even say "claims" in the first place. It's treated as fact, and an unproven fact at that.

--Urthogie 15:47, 24 December 2006 (UTC)

Strongly agree with Urthogie. Claims should be attributed, instead of stated as fact. --JereKrischel 05:14, 26 December 2006 (UTC)

Black-White IQ differences remain when controlling for occupation, socioeconomic level etc.

Claim: "In the United States, Black-White IQ differences of varying magnitudes have been observed when controlling for age (above 3 years), occupation, socioeconomic level, and region of the country."
Source: Arthur Jensen's The G Factor: the Science of Mental Ability
Reason: Where in that cited article is there data that controls for all of the above listed factors with an adequate sample-space that is representative of the world?

--Urthogie 17:50, 23 December 2006 (UTC)

The article regards testing in the U.S. Any caveats about the testing methods can be added, as long as it's not original research. (citable statements regarding their particular test methods)-- Mytwocents 04:21, 24 December 2006 (UTC)

Ok, I recognize I made a mistake in skipping over the "In the United States" clause, but please prove to me that the study controls for age, occupation, and socioeconomic level, and region.
This claim needs to be attributed to Jensen if it's kept.

--Urthogie 15:49, 24 December 2006 (UTC)

Strongly agree with Urthogie. Claims should be attributed, instead of stated as fact. --JereKrischel 05:15, 26 December 2006 (UTC)

Strongly agree with Mytwocents -- Jensen is summarizing a wide range of studies when he makes this claim. This is the kind of 2ndary source that is of the highest value/priority for WP article writing, as it places all of the work of snythesis on the source. No inline attribution is required because the veracity of a 2ndary source should only be questioned by other 2ndary sources -- and no doubt of this claims appears in the literature AFAIK. --W. D. Hamilton 21:24, 26 December 2006 (UTC)

Summaries surely are valuable to Wikipedia writing, but stating one man's summary as fact is not. What part of Wikipedia policy says that summaries provided by 2ndary sources don't have to be attributed?--Urthogie

There is no reason to attribute to Jensen (in line) the belief that claim x is a common belief. Jensen's writings in this context are like a text book. We would not in-line attribute common beliefs reported in textbooks to the authors of those text books. --W. D. Hamilton 22:00, 27 December 2006 (UTC)

Socioeconomic success (SES) has been individually controlled for, and the gap remains even then--yes. I'd like to see proof, though, of a single study that controls for SES, age, occupation, and region all at once. I think that controlling for SES keeping the gap is common knowledge in the field, yes, but I don't know of a study where there has been a control for the four above listed factors all at once. It might exist, but I don't know of it. If it's common knowledge, finding me a source shouldn't take more than a minute.--Urthogie 01:48, 28 December 2006 (UTC)

I'd like to see proof, though, of a single study that controls for SES, age, occupation, and region all at once. -- That's not the claim. Rather, each has been examined individually. I'm actually quite certain that studies have been done in which an enormous battery of SES variables are used as controls. Each additional control (if it is correlated with IQ) will further diminish the IQ gap simply because it is creating a more precise proxy variable for IQ. However, this is beyond the scope of Jensen's argument. He was discussing the ubiquity of the IQ gap, and the fact that no simple differences in social and economic variable can account for it. The APA report has an exensive discussion of SES, which includes an argument for why SES should be insufficient to account for the gap -- this argument places at least part of the explanation on cultural factors. --W. D. Hamilton 17:39, 1 January 2007 (UTC)

I see what you're saying. The thing is, though, that the sentence is written in such a way that its too vague for the reader to know if the variables are controlled for in conjunction or individually, as is the case. Perhaps better than this would be:

None of the following factors completely account for the Black-White IQ gap: age (above 3 years), occupation, socioeconomic level, and region of the country.

--Urthogie 23:29, 1 January 2007 (UTC)

hmm, that sentence was exensively rewritten by an editor who didn't consult the original source. i'll try to dig out to original sentence. --W. D. Hamilton 16:13, 2 January 2007 (UTC)

fixed --W. D. Hamilton 16:32, 2 January 2007 (UTC)

The new sentence you added makes it look like they've done a study on every occupation, which they haven't. My sentence avoids this problem.--Urthogie 19:18, 2 January 2007 (UTC)

My sentence avoids this problem. - Which sentence? --W. D. Hamilton 20:11, 2 January 2007 (UTC)

None of the following factors completely account for the Black-White IQ gap: age (above 3 years), occupation, socioeconomic level, and region of the country.

--Urthogie 02:08, 3 January 2007 (UTC)

The connotation is false. Those factors have very little ability to account for the black white gap (e.g. the gap is 12-15 points when SES is controlled). I added the word "tested" to account for the "every" problem. --W. D. Hamilton 03:41, 3 January 2007 (UTC)

Good.--Urthogie 20:40, 3 January 2007 (UTC)

regression to differing means for different races

Claim(s): "Studies of US comparisons of both parents to children and siblings to each other finding regression to differing means for different races (85 for Blacks and 100 for Whites) across the entire range of IQs, despite the fact that siblings are matched for shared environment and genetic heritage, with regression unaffected by family socioeconomic status and generation examined"
Problematic source:http://www.lrainc.com/swtaboo/taboos/cmurraybga0799.pdf
Reason: The cited study is based on "g-loaded" military intelligence tests, but the article claims this represents IQ.

--Urthogie 18:00, 23 December 2006 (UTC)

If you think the 'G-Loaded' tests are questionable, add a sentence from a reputable source that questions the validity of such tests. -- Mytwocents`

I'm just saying that they're not IQ tests, as the claim says. There are already sources that criticize the military tests. I'll add them if a modified version of this claim is added which says "across the entire range of scores on military intelligence tests."--Urthogie 15:55, 24 December 2006 (UTC)

the military test in question is universally recognized as being a de facto IQ test that is as good at measuring g (what IQ tests are meant to measure) as the best IQ tests. --W. D. Hamilton 21:26, 26 December 2006 (UTC)

It's not universally recognized as such. See the below sources under removed claim #4 for two links that criticize the tests relation to IQ. Secondly, the methodology applied to the test is also criticized in at least on of those links below.--Urthogie 21:38, 26 December 2006 (UTC)

I believe this is covered below. I know of no peer reviewed sources which seriously criticize the armed forces test as a measure of g outside of the context of book reviews of TBC. --W. D. Hamilton 22:02, 27 December 2006 (UTC)

You want a peer reviewed source? How about Hunt, Earl (1999) The Modifiability of Intelligence, Psycoloquy: 10,#72 Intelligence G Factor (14)? It criticizes the military tests as being correlated largely to schooling and not just intelligence:

The case has been best made for the data from the National Longitudinal Study of Youth (NLSY), where it has been shown that an extra year of schooling does influence scores on the Armed Forces Vocational Aptitude Battery (ASVAB) (Cawley et al., 1997). This is hardly surprising to anyone who knows what is on the ASVAB, a lot of its content is related to school subjects. Because Jensen is willing to commit himself to the ASVAB, and to the NLSY data set, when it suits his purposes, he must deal with the schooling effects demonstrated for this test and data set.

This is solid proof that the level of g represented by the military tests has been observed to be limited by two sources: Cawley (primary) and Hunt (secondary). Both peer reviewed, and both criticizing the armed forces test as a measure of g.--Urthogie 01:35, 28 December 2006 (UTC)

(1) You have misread Hunt. He is talking about the malleability of IQ, not the suitability of the the NLSY test as a measure of intelligence. Claiming that it is "modified" by education does not make it not a measure of intelligence. IQ and educational attainment are correlated around .5, and thus there is a substantial relationship between the two. The psychometric properties of the NLSY test have been well studied and it is a substantial measure of g. The relationship of the test to education was part of the analysis in TBC (p. 589-592). Suffice it to say, the claims made about the NLSY can and have been made about many tests of cognitive ability, where some experts claim that education boosts IQ. This claim does nothing to change the fact that the NLSY test is treated by experts as a measure of intelligence.

My original point was that the armed forces tests measure education as well as intelligence. Hunt and Cawley confirm this. I never said that being ""modified" by education does not make it not a measure of intelligence." I said that has "been observed to be limited" in its measurement of IQ and, in turn, intelligence. So it seems like you're misreading me, rather than me misreading Hunt/Cawley. Secondly, which IQ test is regarded by experts as the best is a somewhat loaded question. Certain IQ tests are biased by education, others less by education and more by culture, so they'll be better for different things. The question is not simply quantitative, it is also qualitiative. "--Urthogie 23:40, 1 January 2007 (UTC)

(2) The regression phenomena has been studied widely. The TBC study is merely the largest and most representative of the U.S. population. Other studies include Jensen (1973, pp. 107-119; looking at siblings) and (Jensen, 1998, p. 358 WRT parents). Thus, the regression phenomena is supported outside of TBC data, which is simply the largest and most accessible source. --W. D. Hamilton 17:53, 1 January 2007 (UTC)

Use other sources if they're better. We'll see if they are.--Urthogie 23:40, 1 January 2007 (UTC)

Those sources were already cited. NLSY data was cited for the 2nd half of the sentence where SES and generations were added to the regression as controls. --W. D. Hamilton 16:35, 2 January 2007 (UTC)

NLSY is not enough to warrant inclusion in the graph, as it doesn't control for education. Education is a major factor, and by ignoring its effect, we're misinforming the readers.--Urthogie 19:21, 2 January 2007 (UTC)

Rhetorical questions -- Who says that this should have been done? How would you control for education in a study of IQ given that in at least one study 66% of variation in educational outcomes can be explained by a prospective measure of g. I believe these are your own ideas and thus constitute original research. Moreover, they aim at a different level of analysis than what is being pursued in the regression experiments. There's nothing about the NLSY and education that doesn't apply to any IQ test. A look at the WAIS, for example, would show that it is very similar in content to the AFQT. --W. D. Hamilton 19:31, 2 January 2007 (UTC)

These aren't my original ideas. I get them from secondary sources. And even if 66% is correct (or at least so says that study), 100 minus 66 is 34. A significant margin of error, even assuming that study you handpicked is correct.--Urthogie 19:44, 2 January 2007 (UTC)

Are we still talking about -- The cited study is based on "g-loaded" military intelligence tests, but the article claims this represents IQ. -- because I think we've got it covered that the AFQT is as much an "IQ test" as tests which are commonly called "IQ tests" (e.g. WAIS, Raven's) and they all show strong correlations with educational attainment? --W. D. Hamilton 20:10, 2 January 2007 (UTC)

There are several reasons why it's important not to assume this test = IQ. Here are some obvious ones:

The differences between various IQ tests are not just quantitative but also qualitative.
Our readers have a right to know what test the research is based on. (There are several reasons for this I can get into if you like.)
Assuming this test = IQ means accepting as fact the research that proves it to be so. Scientific research produces strong theories, but not facts.
Even if it was fact (which it isn't) that this IQ test was the best, or equal to the best, it would still not give it safety from criticism and analysis by secondary sources.

I could probably list more reasons if I thought harder about it, but I think you get the point...-Urthogie 02:07, 3 January 2007 (UTC)

I believe your argument assumes that there is a definition of "IQ test" which has certain necessary and sufficient conditions and that certain tests either satisfy or not these conditions. As described, this is not the case. (There's certainly no problem with including notes about which particular tests were used in particular studies if there's some reason for that to be relevant, but most conclusions are either drawn from a mix of studies with a mix of tests or the particular test used is a detail buried in the text of the original sources.) However, there is no test called "IQ test". Rather there are dozens of tests which aim to measure cognitive ability and yet other tests that measure cognitive ability equally or nearly equally as well despite being designed with more specific purposes in mind (the SAT and AFQT are in this class). These tests are known as IQ, but IQ itself is not a definite thing, and the "IQ" score that each tests measure is actually a slightly different mix of psychometric variables, but all measure mostly g. The convention of referring to "IQ scores" or just "IQ" is described in the intelligence section: As commonly used, "IQ test" denotes any test of cognitive ability, and "IQ" is used as shorthand for scores on tests of cognitive ability. The psychometric ground work to establish various tests as measures of g is beyond the scope of this article as it does not figure prominently in any discussion of "race and intelligence" that I have read. --W. D. Hamilton 03:36, 3 January 2007 (UTC)

You're misreading my argument. I don't have a problem with calling this an IQ test. I have a problem with claiming its results (or the result of any IQ test) are completely interchangable with an accurate, representative IQ (intelligence quotient). IQ tests are meant to return a completely accurate IQ (intelligence quotient), but that doesn't mean they completely do.--Urthogie 20:43, 3 January 2007 (UTC)

You're suggesting that we can't write about "IQ tests" in the abstract because no single "IQ test" is an IQ test of perfect form? That's not a meaningful notion for psychometric reasons[1], but more importantly it fails for practical reasons. No thermometer is a perfect measure of temperature, but we can talk about thermometer readings in the abstract w/o mentioning the design of the device. Moreover, the (rather high) reliability of IQ tests is itself a factor which is typically corrected for in psychometric studies.

[1] IQ isn't a single thing, but rather it's a vehicle -- class of vehicles -- to measure the construct g among others -- from the article: As commonly used, "IQ test" denotes any test of cognitive ability, and "IQ" is used as shorthand for scores on tests of cognitive ability. -- see also IQ#IQ_and_general_intelligence_factor --W. D. Hamilton 06:11, 4 January 2007 (UTC)

A thermometer isn't perfect, but its margin of error is perfectly detectable. The only factor it measures is heat-- the vibration of atoms. So we can say on that day it was X temperature with X margin of error. You can't do that with IQ tests and IQ, as there are infinite number of variables effecting the resulting scores, and the quality of the tests. The definition of IQ isn't "What's measured by IQ"-- the definition of IQ is an accurate quotient/reading of one's intelligence.--Urthogie 18:56, 4 January 2007 (UTC)

That doesn't address my points. Moreover, the reliability of IQ tests (their retest reliability) and their other psychometric properties are well studied. They too have empirically defined margins of error. To the extent that I can imagine how we might implement your suggestion, I can't imagine it anything but undue confusion for the reader. --W. D. Hamilton 19:33, 4 January 2007 (UTC)

It's not confusing to hear the results of a test. It's misleading to hear IQ in place of that. With thermometers there is no scientific secondary source claiming that the margin of error ascribed to them is even 1% off--the margin of error is basically known. With IQ tests it's different-- there is scientific debate as to what percent off they are.--Urthogie 19:38, 4 January 2007 (UTC)

That's simply not true. These tests are mostly commercial products. The companies that make them ascertain their reliability and validity in a variety of ways. IQ tests (in general) are the most reliable and validity instruments in psychology, being some of the oldest. --W. D. Hamilton 19:49, 4 January 2007 (UTC)

It's not confusing to hear the results of a test. It's misleading to hear IQ in place of that. -- Are you talking about reporting "raw scores" instead of "normalized scores" for *all* data reported in this article? First, it is a direct algebraic conversion from one to the other. Second, most results are based on an analysis of multiple individual studies, and so particular results are not important. WP articles are encyclopedia articles, not technical reports or novel literature reviews. It's reporting, not analysis, and so the form in which results are published is the form in which we should report them. --W. D. Hamilton 20:06, 4 January 2007 (UTC)

Lastly (I hope) unreliability attenuates correlations, it doesn't enhance them. --W. D. Hamilton 20:19, 4 January 2007 (UTC)

poverty rate 25% lower

Claim: "For example, a randomly selected group of Americans with an average IQ of 103 had a poverty rate 25% lower than a group with an average IQ of 100. Similar substantial correlations in high school drop-out rates, crime rates, and other outcomes have been measured.
Reference/source of claim: For this calculation, Herrnstein and Murray alter the mean IQ (100) of the U.S. National Longitudinal Survey of Youth's population sample by randomly deleting individuals below an IQ of 103 until the population mean reaches 103. This calculation was conducted twice and averaged together to avoid error from the random selection.(Herrnstein_and_Murray 1994, pp. 364-368) Discussed further in the section #Significance of group IQ differences #Within societies.
Why I removed this claim:

This controversial claim should not be in the lead. The lead is a place to lay out the established facts, not to debate and weight a specific study.
This study shouldn't be referenced in this article without a discussion of the flaws that have been claimed to exist in its methodology.
This study shouldn't be referenced in the article without explicitly saying who conducted it (the authors of the bell curve). Putting it simply as a reference is not enough, especially for such a controversial subject as this.
I found the following criticisms of this finding over the course of 15 minutes:

First, consider the definitions of the two key variables used in the authors' empirical study: IQ and family background. In their empirical analysis, the operational definition of IQ is the Armed Forces Qualifying Test (AFQT) score, which is based on an aggregation of a subset of 10 separate exams given to more than 12,000 youth in 1980. The youth were 15 to 23 years old when the tests were administered. These tests were designed to predict success in military training schools, and there is much evidence that they do so, although by no means are they 100-percent reliable. Most of these tests appear to be achievement tests rather than ability tests (i.e., they partly measure factual knowledge and not pure ability). A subset of four tests was assembled to define the Armed Forces Qualifying Test (AFQT) score used by Murray and Herrnstein. Ironically, the authors delete from their composite AFQT score a timed test of numerical operations because it is not highly correlated with the other tests. Yet it is well known that in the data they use, this subtest is the single best predictor of earnings of all the AFQT test components. The fact that many of the subtests are only weakly correlated with each other, and that the best predictor of earnings is only weakly correlated with their "g-loaded" score, only heightens doubts that a single-ability model is a satisfactory description of human intelligence. It also drives home the point that the "g-loading" so strongly emphasized by Murray and Herrnstein measures only agreement among tests--not predictive power for socioeconomic outcomes. By the same token, one could also argue that the authors have biased their empirical analysis against the conclusions they obtain by disregarding the test with the greatest predictive power.
— Nobel Prize Laureate in Economics, professor James heckman

What Herrnstein and Murray used to measure IQ is actually a measure of education as well as intelligence. All the people tracked in the National Longitudinal Study of Youth took the Armed Forces Qualifying Test, which Herrnstein and Murray treat as a good measure of intelligence. Because the material covered in the test includes subjects like trigonometry, many academic critics of The Bell Curve have objected to its use as a measure only of IQ and not at all of academic achievement. Herrnstein and Murray concede in the footnotes that scores tend to rise with the subjects' education--but they seriously underestimate the magnitude of this rise, as Case Study Three shows. And they resist the obvious inference that the test scores are measuring something other than intelligence.
Most of The Bell Curve's analysis is devoted to proving that IQ has more predictive power than parental "socio-economic status." But Herrnstein and Murray's method of figuring socioeconomic status seems designed to low-ball its influence, as Case Study Four explains.
Herrnstein and Murray begin their discussion of the National Longitudinal Study of Youth data by announcing that they aren't going to analyze the effect of education, because education is too much a result of IQ. It's not an independent variable. (Of course, according to their theory, socioeconomic status is also a result of IQ, but somehow, that doesn't stop them.)
— Professor and Dean of Journalism Nicholas Lemann

So, in conclusion, this study is worth mentioning, but not in the lead, as the credibility of its methodology has been numerously challenged (this is just two sources.). Please argue against my point before reverting me. I'll give you a week between replies, and we can keep it as it was until the debate is over. Thanks for keeping cool and discussing this, --Urthogie 21:00, 26 December 2006 (UTC)

in the context given, the particular values found in TBC are of no value; they are merely examples of the kind of social-economic factors that are correlated with IQ and thus would be different in two populations with different IQ all else being equal (because they are tail phenomena). --W. D. Hamilton 21:28, 26 December 2006 (UTC)

Actually, in the context-given, all that matters is TBC-- it's the reference used to support the claim.--Urthogie 21:40, 26 December 2006 (UTC)

You have it exactly backwards. The topic of this article is not TBC and its critics. Moreover, your criticisms miss the point that none of the claims made in TBC about the practical validity of IQ are novel to TBC. They have been extensively studied since the 70s, and the APA report substantiates as accepted all of the relevant points from TBC cited in this article. --W. D. Hamilton 21:58, 27 December 2006 (UTC)

The question isn't the entire article, it's the specific claim here. The claim is based on criticized data (TBC), and yet its findings are presented as fact. As far as the APA report, they do note the affect that IQ has on socioeconomic success, however they also add:

This effect, in turn, is substantially mediated by education: the brother with the higher test scores is likely to get more schooling, and hence to be better credentialled as he enters the workplace.

TBC, of course, chose random samples of people with IQ's 100 and 103-- they refused to control for education, and TBC's critics rightfully point out that this causes problems in the tests as well ("What Herrnstein and Murray used to measure IQ is actually a measure of education as well as intelligence." --Lemann). And the APA report seems to back up their position. --Urthogie 00:55, 28 December 2006 (UTC)

TBC claim is the magic wand version of raising IQ. It does not specify how it is achieved, and instead assumes that everything else remains the same. The point being made in TBC and here is that small differences in IQ account for large differences in life outcomes. How these differences are causally related is not specified here or in the context of TBC and is moot for this point. The larger point of IQ correlating with life outcomes is, as I pointed out, among the conclusions of the APA report, which provides evidence that it is a widely accepted finding. The WSJ statement further supports this view. There is no criticism of TBC or other sources that will make this not a substantiated majority conclusion. --W. D. Hamilton 17:57, 1 January 2007 (UTC)

I'm not arguing that there isn't a correlation. I'm arguing that the correlation has been shown to be very spurious if you look at the data and methods that led to it. I base my argument on secondary sources. AS far as the APA, they go out of their way to show how the reasoning that gets such statistics and claims that they're meaningful (not specious) as this is flawed. They go out of their way to acknowledge the big effect that education has-- they're not so unscientific as to condone such garbage as this claim. This statistic is extremely misleading, and should not be included except as a demonstration of how ignoring education distorts statistics.--Urthogie 23:49, 1 January 2007 (UTC)

(1) Education was not ignored by TBC. (2) The criticism regarding IQ/education is not specific to the NLSY test, but rather a general (AFAIK minority) criticism of IQ tests. The general issue is whether (a) IQ is merely a measure of education or whether (b) IQ causes educational outcomes to a large extent. The former (a), predicts that the correlation between life outcomes and IQ is merely mediated by education and thus IQ is a synonym for education. The later hypothesis (b) predicts that educational attainment is largely a consequence of differences in IQ. (3) However, the answer to this question does not matter for the particular calculation presented. The causal connections between the variables shown are not assumed to be in any particular direction. Rather, the point is to demonstrate that small differences in IQ are associated with large differences in some life outcomes. This is true regardless of whether IQ is simply a reflection of education, or an independent causal factor. (For more on this topic, see the latest issue of Intelligence for three papers. Conclusion from one of the papers: "the present study provides evidence that psychometric intelligence is predictive of future achievement whereas achievement is not predictive of future psychometric intelligence. This temporal precedence is consistent with the theoretical position of Jensen (2000) that intelligence bears a causal relationship to achievement and not the other way around."

In summary, the conclusion that the statistics are "misleading" are due to your misreading of the sources you presented and a confusion of levels of analysis. --W. D. Hamilton 16:11, 2 January 2007 (UTC)

You said: However, the answer to this question does not matter for the particular calculation presented. Actually, the significance of education does matter. That's why it leads to such ridiculous and misleading claims as this one. If we keep this claim, it should be sourced and the criticism that it doesn't control for education needs to be added. Also, it should be removed from the lead.--Urthogie 19:25, 2 January 2007 (UTC)

The specific criticism WRT education and this kind of calculation needs to be sourced; the calculation is sourced and accurate. My claim about education not mattering is perfectly correct. Under the "synonym" theory of IQ/education, IQ is identical to educational attainment, and so all that can be criticized is that education could/should be perfectly substituted in this calculation. However, in a discussion of "intelligence" there is an obvious assumption that it is okay to talk in terms of intelligence, which is what we are doing in this article. --W. D. Hamilton 19:37, 2 January 2007 (UTC)

The synonym theory is not fact, it's theory. Our definition of intelligence is not based on the synonym theory, but on all the research in this field.--Urthogie 19:47, 2 January 2007 (UTC)

I believe we've become confused. --W. D. Hamilton 20:07, 2 January 2007 (UTC)

Whether the material is appropriate for the lead is a different question about which I have less strong feelings. --W. D. Hamilton 19:37, 2 January 2007 (UTC)

So we'll (re)move it from the lead.--Urthogie 19:47, 2 January 2007 (UTC)

A more interesting example might be better, but an example is always nice to ground it. --W. D. Hamilton 20:07, 2 January 2007 (UTC)

I read a lot of encyclopedias, and I'm yet to find one that would cover a controversial subject with a controversial example in the lead. More likely it would instread try quoting the APA saying that IQ has significant effects, or something along those lines.--Urthogie

conclusions about tests of cognitive ability are now largely accepted

Claim: Several conclusions about tests of cognitive ability are now largely accepted:

For people living in the prevailing conditions of the developed world, cognitive ability is substantially heritable, and while the impact of family environment on the IQ of children is substantial, after adolescence this effect becomes difficult to detect.

Source(s): None.
Why I removed the claim: This is not largely accepted. Robert J. Sternberg, Elena L. Grigorenko, and Kenneth K. Kidd of Yale University wrote an article for the American Psychological Association (APA) which criticized this idea:

heritability is itself a troubled concept. Are
differences in intelligence between so-called races heritable? This question is difficult to answer in part because it is difficult even to say what can be concluded from the heritability statistic commonly used. Consider some facts
about heritability (Sternberg & Grigorenko, 1999)...

It goes on for several more pages discussing the problematic nature of studies of heredity.

I'm not against citing hereditarian interpretations of the data. I'm just saying that we shouldn't claim such conclusions are "largely accepted." Thanks, --Urthogie 03:17, 27 December 2006 (UTC)

The citation of Sternberg et al does not contradict the claims that IQ is heritable, but more importantly it does not contract the claim that the heritability and genetic structure of IQ is "largely accepted". The APA report establishes these findings as largely accepted as do the other consensus reports. --W. D. Hamilton 21:56, 27 December 2006 (UTC)

First off, Sternberg helped write that report with 10 others, so if the report defines whats "largely accepted", then so does Sternberg to a certain extent. Secondly, the report was released in 1995 (edited version published in 1996), while Sternberg & Grigorenko (which is cited several times in agreement by other scholars in the field) was 1999. Thirdly, Sternberg was president of the APA in 2003. He is by no means representitive of a minority view.--Urthogie 02:24, 28 December 2006 (UTC)

Sternberg is quoted (in this article) to the extent that he holds a "minority" view regarding hereditarianism. Again, read the APA report and/or the major manuals/textbooks/encylopedias on intelligence (published since 1996 of which Sternberg is an editor of some). --W. D. Hamilton 23:03, 28 December 2006 (UTC)

Telling me to read more doesn't help your point at all. This isn't a "largely accepted" view. If it was, you could demonstrate it as such. But instead of demonstrating it, you tell me to read more. Wikipedia:Reliable sources says it's better to remove possibly false and misinformative claims than to keep them there. Unless you can demonstrate that it's a "largely accepted" view, we shouldn't present it as such, as we risk misinforming people.--Urthogie 23:38, 28 December 2006 (UTC)

I have pointed to the sources which directly indicate that these views are largely accepted. I will enumerate them here, at which point it falls to you to demonstrate that their claims of representing the majority view are incorrect. Simply citing a member of the minority view (self described) does not constitute evidence against the claim of what the majority view is.

These are the sources which specifically report what is the majority view:

the WSJ statement - Mainstream Science on Intelligence PDF
the APA report - Intelligence Knowns and Unknowns
the Snyderman and Rotham (1987) survey of "intelligence experts"

The other sources I mentioned (the major manuals/textbooks/encyclopedias on intelligence) as a whole also constitute a representation of the majority view, but they require further synthesis than the three cited above, which directly claim to represent what the majority view is. A reading of these works establishes the majority conclusion that h^2 is very high and c^2 approaches zero in adulthood. There is no WP:RS issue here. --W. D. Hamilton 17:34, 1 January 2007 (UTC)

Ok, nice work with the sources-- I'll admit that withing the field of intelligence research (a qualifier that needs to be added to the sentence) Sternberg represents a minority view on heredity-- I was wrong in this regard. Here's what I'll say, though: the entire article is framed around what is originally listed as "largely accepted", so to ignore a minority viewpoint is to essentially exclude it from the rest of the article. This is why there is no real discussion of the suggested problematic nature of heredity as put forward by Sternberg et al. Instead of discussing what is "largely accepted" and basing the entire article on that, we should start the article with what is universally accepted among scientific, peer reviewd intelligence researchers, so that the article can go on to discuss all their work and debates, not just the majority views.--Urthogie 00:00, 2 January 2007 (UTC)

Even Sternberg's caveat's about heritability (as I recall them) don't directly counter the text of the claim made here. If there were some particular point about race and intelligence [1] that warranted mention of his opinions, then we should surely include them. However, it is undoubtedly impossible to include all tangentially related minority views about race or intelligence.

[1] Flynn's "Heritability Estimates Versus Large Environmental Effects: The IQ Paradox Resolved" paper is one such example that is given lengthy description in the environmental explanations section. A comparison of that paper and Sternberg's comments would reveal a lot of similarities I suspect. --W. D. Hamilton 15:52, 2 January 2007 (UTC)

I would rest this point if we'd add a subsection discussion of Heritability to the background information section. By the way, I commend you for proving your point to someone who disagrees with you here.--Urthogie 19:30, 2 January 2007 (UTC)

What would we say about it? --W. D. Hamilton 19:38, 2 January 2007 (UTC)

That the mainstream consensus is that it's not so troubled a concept and that its effect increases with age, etc. Also, we'd note the minority opinions of Sternberg and also some remarks by non-intelligence researchers who happen to discuss heritability in their scientific work.--Urthogie 19:49, 2 January 2007 (UTC)

Okay if its related. I think we covered increase with age. --W. D. Hamilton 20:06, 2 January 2007 (UTC)

Should we remove the section 'Employment tests and school achievement' ?

What proof is there that the SAT and GRE correlate to intelligence? If there's only interpretation, and not any study which demonstrates the correlation, then why do we even have this section? Sure, it seems like a good statistic to consider, but if it's not proven to be related to 'g' it seems irrelevant to the article. I'll remove it in a week if there's no disagreement/dialogue. Thanks, --Urthogie 03:44, 27 December 2006 (UTC)

Several studies have looked at the correlation between SAT and IQ (e.g. work by Detterman et al.; but also Roth et al 2001 for a massive meta-analysis). However, on its own the "achievement" gap is of comparable importance to the "IQ" gap in that these tests are used as measures of cognitive ability by schools. --W. D. Hamilton 21:54, 27 December 2006 (UTC)

Frey & Detterman (2004) said:

There is little evidence showing the relationship between the Scholastic Assessment Test (SAT) and g (general intelligence).

Let's take Detterman and Frey at their word and assume that as of 2004 nothing was proven as far as SAT's correlating to g. If that's the case, then all we seem to have attempting to prove this is Detterman and Frey. However, Bridgeman handed them their asses in Bridgeman 2005. Detterman issued a rejoinder, yes, but all that means is we have a criticized minority opinion.. unless you know of other studies since 2004...? If you don't know of other studies in support of Detterman, I think it's misleading to give an entire section devoted to what only Detterman and Frey think is valuable in representing g.--Urthogie 03:12, 28 December 2006 (UTC)

Until 2004' there had been no large scale studies of the SAT, but the objections of Bridgeman 2005 are limited only to the regresion equations and make no challenges against the claim that the SAT is correlated with g. Odd that you misunderstood this. However, the finding has been replicated in part by Rushton and colleagues in their study of g differences between males and females using the SAT -- they find that the SAT measures a general factor g.

More later... --W. D. Hamilton 15:43, 28 December 2006 (UTC)

Wasn't the study that tried to show the correlation between SAT's and g based on regression equations? If the regression equations are off, then the data is off, right? Correct me if I'm wrong (I may be on this point). Also, please provide a link to Rushton.--Urthogie 19:35, 28 December 2006 (UTC)

I'm away from home and so it would take some time to find the Rushton piece: check PAID and Intelligence. The criticism paper argues that the regression equations predict an unreasonable average IQ for an average SAT score. However, this is outside of the SAT/IQ range of the training set on which the regression equations were built, and thus is an example of extrapolating outside of the data. The 2004 study actually produced two regression equations which yield somewhat different predictions. This is simply a limitation of the available data -- more data is needed to produce better regression equations. However, regression and correlation are not identical, and the point of interest is merely that the SAT measures cognitive ability in a similar fashion to an IQ test. This is a finding that should come as little surprise given the form and history of the tests, and the fact that g provides essentially all of the predicitive ability of cognitive ability tests (and the SAT predicts educational outcomes). Further support for the SAT/g relationship comes from the 2006 ISIR conference -- one talk looked at reducing the length of the SAT while maintaining the predictive ability by using only the most g loaded items. --W. D. Hamilton 22:59, 28 December 2006 (UTC)

"However, regression and correlation are not identical, and the point of interest is merely that the SAT measures cognitive ability in a similar fashion to an IQ test." Correct me if I'm wrong, but regression is required to prove the correlation is not specious. So if the regression is off, shouldn't the supposed correlation hypothesis also be flawed? As far as Rushton and the ISIR conference, I'll give you a week to produce links to those sources. I can almost guarantee you that anyone who does a study showing that SAT has high correlation to g has made some error or another (if I'm proven wrong, of course, I'll admit it). I've taken the SAT's, and I did amazingly well-- on the sections I read about online in advance. The new writing section fucked up a lot of people because you get a perfect score only if you write it in their arbitrary format. This is why I'm confident when I say you won't find any solid proof that it correlates with g. Anyways, good luck on that task, I'm giving you or anyone else a week before any edits are made in this regard.--Urthogie 23:46, 28 December 2006 (UTC)

The answer to your first question (and my comment itself) comes from Frey and Detterman. You can read the papers to your own satisfaction. In short, the regression equations work for the range of values tested, but there's not enough breadth of IQ scores among SAT takers to accurately predict the SAT/IQ scores of less intelligent individuals. This does nothing to change the fact that SAT and IQ scores are correlated among those who take the tests. No other citations are needed at that point, but I assumed you could find more if you were interested. The Rushton citation is:

Douglas N. Jackson and J. Philippe Rushton, Males have greater g: Sex differences in general mental ability from 100,000 17- to 18-year-olds on the Scholastic Assessment Test, Intelligence, Volume 34, Issue 5, , September-October 2006, Pages 479-486. [1] -- "We found (1) the g factor underlies both the SAT Verbal (SAT-V) and the SAT Mathematics (SAT-M) scales with the congruence between these components greater than 0.90; (2) the g components predict undergraduate grades better than do the traditionally used SAT-V and SAT-M scales; (3) the male and the female g factors are congruent in excess of .99;"

A more direct response to your title question is that we should not remove discussion of employment and IQ tests because scholars in the field treat them as tests of cognitive ability. One of the most comprehensive studies of group difference in intelligence looked exactly at that question: Roth, P. L., Bevier, C. A., Bobko, P., Switzer, F. S. and Tyler, P. (Sum 2001). "Ethnic Group Differences in Cognitive Ability in Employment and Educational Settings: A Meta-Analysis". Personnel Psychology 54 (2): 297-330. --W. D. Hamilton 17:04, 1 January 2007 (UTC)

Addendum: it is not my aim to convince you that the SAT measures g, but rather to convince you that the scientific literature (from many sources) includes the unchallenged claim that the SAT measures g (as per above and as F&D point out, the criticism of there work does not dispute this). --W. D. Hamilton 17:23, 1 January 2007 (UTC)

It seems like no scientific studies or analyses have confirmed or denied Rushton's findings as of yet. Do Rushton's findings take into account any of the following when making the claim that IQ correlates highly to G:

education
whether the person has studied in advance
effort put into the test

If not, we should list these and any other factors that Rushton's study do not control for. If we do choose to keep this section, it's important that the section explicity state in the text that only these two studies have found this correlation, (rather than saying weasilily that "gaps are seen") and that major flaws have been found in the first, and that the second one has as of yet not been re-validated or rebutted by other studies, because it was published earlier this year.--Urthogie 01:02, 2 January 2007 (UTC)

That seems like original research to me. Somewhere in the literature there may be a related reference to hang that claim on, but it would surely require a citation. --W. D. Hamilton 15:54, 2 January 2007 (UTC)

How is it original research? It's simply stating facts about a secondary source so as to inform our readers.--Urthogie 19:31, 2 January 2007 (UTC)

Implicit in the claim is the assumption that such factors should be taken into account, a view point which needs to be supported by citation. As suggested, it constitutes a novel analysis/comment on the data. Which raises the question as to why those particular factors and not others. A similar explication about what factors aren't controlled for can be made about a great many of the studies described in this article, and so there needs to be a specific reason for making such comments. --W. D. Hamilton 19:41, 2 January 2007 (UTC)

What I'm saying is that any time a reliable secondary source brings up an issue or an assumption, it can be discussed in the article and sourced to them. In this case, I've already provided such sources that demonstrate how education has an effect on IQ tests of all sorts. Your suggestion is we just discuss this effect in the definition section for Intelligence. I disgaree with this approach, as it ignores the fact that different tests will be effected to different degrees by education.--Urthogie 19:54, 2 January 2007 (UTC)

What sources say "different tests will be effected to different degrees by education"? --W. D. Hamilton 20:05, 2 January 2007 (UTC)

In this particular case, it would be odd to control for education when looking at the ability of IQ to predict measures of scholastic aptitude, which can itself be taken as a measure of education. --W. D. Hamilton 19:45, 2 January 2007 (UTC)

This oddity can be noted in the article.--Urthogie 19:54, 2 January 2007 (UTC)

How many steps removed from the original paper would be be at that point? The critics of Frey and Detterman did not mention controlling for education AFAIK -- so why should we? --W. D. Hamilton 20:05, 2 January 2007 (UTC)

It's a sign of a good encyclopedia when it makes an effort to make clear what a given study accomplished using the criteria of a field. It only becomes original research when we start producing our own findings.--Urthogie 20:08, 2 January 2007 (UTC)

That's not my reading of WP:NOR at all. Effectively, it limits us to doing what a reporter could do and little else. --W. D. Hamilton 20:51, 2 January 2007 (UTC)

I just read over WP:NOR and it seems like it's not very clear cut here. Am I correct to say that you're arguing that what I'm suggesting would fall under the heading "Synthesis of published material serving to advance a position"? Specifically, which sections of WP:NOR do you think would be violated?--Urthogie 01:45, 3 January 2007 (UTC)

I believe it would be an example of: It introduces an analysis or synthesis of established facts, ideas, opinions, or arguments in a way that builds a particular case ~~favored by the editor,~~ without attributing that analysis or synthesis to a reputable source; Even if it were generally true that people like to note about IQ studies whether they do or do not control for a set of variables, it would be OR to make that point about a particular case. The (plagiarism) example at the NOR page for the quoted rule is very much like that. I struck out the part about "favored by the editor" because intent seems irrelevant. --W. D. Hamilton 03:26, 3 January 2007 (UTC)

Ok, this is arguable original research. I'm gonna play it safe and assume you're right in suggesting so much. However, until other studies are published confirming, denying, or criticizing Rushton's findings, I would ask that the following addendums be made in the section that they appear:

That as of 2004 there were no studies that supplied any possible proof of an SAT and g relation.
That Detterman's findings were rebutted because of their regression methods, and that Detterman issued a Rejoinder.
That aside from Detterman, there is only one other study--Rushton's-- that tests the same thing as him and comes to the same conclusion.
Any studies that have attemped to disprove a correlation between SAT and g should be discussed.
That as of today, no studies/papers have confirmed, denied, or observed flaws in Rushton's findings, because they were published this year.
Rushton's application of g is actually off in its correlation to grades, as girls get better grades than boys. (This is noted by a secondary source, I can find it if you don't want to).

This is what comes to mind. We have to be as careful as possible, within the realm of WP:NOR, when presenting very new research findings such as this one by Rushton.--Urthogie 21:00, 3 January 2007 (UTC)

We're now totally off topic. The title question was Should we remove the section 'Employment tests and school achievement'? The primary reason the answer is "no" has nothing to do with the IQ/GRE/SAT correlation[1], but rather that the gap in employment tests and school achievement are an intimate part of the literature on "race and intelligence". (For example, Nisbett 2005, Flynn and Dickens 2006, Rushton and Jensen 2006, Gottfredson 2005; and that's just off the top of my head from the last two years.) That in and of itself is sufficient reason for why the section should not be removed.

[1]A discussion of the IQ/GRE/SAT relationship may belong somewhere in Wikipedia, but it's not the appropriate topic for an exposition in this article -- especially this summary style top-level article. --W. D. Hamilton 06:05, 4 January 2007 (UTC)

We are off topic, but we're still discussing the section. I disagree with you that the above stuff shouldn't be noted.--Urthogie 19:01, 4 January 2007 (UTC)

I can't see how a the inclusions you suggested can be justified within WP policy; but setting NOR aside, it would seem to be an entirely off-topic addition to describe the technical minutia of the papers that link the SAT to IQ in that section. --W. D. Hamilton 19:30, 4 January 2007 (UTC)

The entire section is only valuable information to the extent that the connection has been proven. So a good amount of time should be added discussing the relation and what has and hasn't been proven as far as that.--Urthogie 19:41, 4 January 2007 (UTC)

That's an interesting opinion, but clearly your own. No published opinion I know of claims that the achievement gap is not important in and of itself (it's a frequent topic of discussion in NY Times and other papers and several books have been written about it). But moreover, there appears to be no disagreement in the related literature that achievement tests are proxies (to some degree) for intelligence. Nisbett (2005) takes them as strong proxies. --W. D. Hamilton 19:52, 4 January 2007 (UTC)

removed training prospects from IQ cohorts graph

The chart said the "training prospects" were based on Wonderlic, which I can only assume to be the Wonderlic Test, which is applied by the NFL and has no scientifically proven correlation to 'g'. Correct me if I'm wrong, and we can revert it and discuss it before further action is taken. Thanks, --Urthogie 04:58, 27 December 2006 (UTC)

The wonderlic is a recognized IQ test which only happens to be popularly known because it is used by the NFL. Vocabulary tests are the best single measures of g, and even a short test has high reliability/validity/heritability. However, the general point of the trainability/employment prospects of people of different IQ levels is independent of the WPT itself. --W. D. Hamilton 21:51, 27 December 2006 (UTC)

What is the source for the training propects, then, if they're independently proven? And why was Wonderlic ever mentioned, then? From what I've been reading just now Wonderlic seems to be 10% off from other IQ tests according to some measures.--Urthogie 03:20, 28 December 2006 (UTC)

First, I will point out that your original reasoning has changed: the Wonderlic Test, which is applied by the NFL and has no scientifically proven correlation to 'g'. Second, the Wonderlic paper is cited because the descriptions of "training prospects" it offers are clear, short, and representative of the findings from other studies. Third, the issue of the WPT itself is moot because the findings of Wonderlic (1992) are for our purposes identical to those of the bulk of other studies (as reflected by the description of the practical validity of IQ in the APA report). The WPT is not at issue here, only the training prospects of individuals with different IQ levels. Here, there is a mountain of data from personnel psychology to support the predictive validity of IQ both within and between occupations [see for example Hunter, J.E. and Hunter, R.F. (1984). Validity and utility of alternate predictors of job performance. Psychological Bulletin, 96(1):72-98.]. --W. D. Hamilton 15:39, 28 December 2006 (UTC)

Here's the APA report weighing in on "training prospects":

In contemporary American society, the amount of schooling that adults complete is also somewhat predictive of their social status. Occupations considered high in prestige (e.g., law, medicine, even corporate business) usually require at least a college degree-16 or more years of education-as a condition of entry. It is partly because intelligence test scores predict years of education so well that they also predict occupational status, and even income to a smaller extent, (Jencks, 1979). Moreover, many occupations can only be entered through professional schools which base their admissions at least partly on test scores: the MCAT, the GMAT, the LSAT, etc. Individual scores on admission-related tests such as these are certainly correlated with scores on tests of intelligence.

and...

In section II we noted that intelligence test scores predict occupational level, not only because some occupations require more intelligence than others but also because admission to many professions depends on test scores in the first place. There can also be an effect in the opposite direction, i.e. workplaces may affect the intelligence of those who work in them. Kohn and Schooler (1973), who interviewed some 3000 men in various occupations (farmers, managers, machinists, porters...), argued that more "complex" jobs produce more "intellectual flexibility" in the individuals who hold them. Although the issue of direction of effects complicates the interpretation of their study, this remains a plausible suggestion.

and...

Using the well-established correlation between intelligence test scores and occupational level, Flynn (1991, p.99) calculated the mean IQ that a hypothetical White group "would have to have" to predict the same proportions of upper-level employment. He found that the occupational success of these Chinese Americans, whose mean IQ was in fact slightly below 100, was what would be expected of a White group with an IQ of almost 120! A similar calculation for Japanese-Americans shows that their level of achievement matched that of Whites averaging 110. These "over-achievements" serve as sharp reminders of the limitations of IQ-based prediction.

and...

Test scores also correlate with measures of accomplishment outside of school, e.g. with adult occupational status. To some extent those correlations result directly from the tests' link with school achievement and from their roles as "gatekeepers." In the United States today, high test scores and grades are prerequisites for entry into many careers and professions. This is not quite the whole story, however: a significant correlation between psychometric intelligence and occupational status remains even when measures of education and family background have been statistically controlled.

So, to summarize: the APA report acknowledges a modest correlation between occupation and test scores. It also notes the correlation may result largely from culture and schooling (which exist as variables even when socioeconomic success is controlled for). It is therefore definitely a hereditarian POV to include the "training prospects" column in the graph, because it give the impression that the training prospects result from the IQ, rather than being correlated to it through largely through culture and schooling, as the APA report suggests. Especially considering that we're talking about a graph here, something that readers will just eye over, often skipping the rest of the article, it's very misleading to present "training prospects" next to IQ. I've taken classes in Visual Communications and they talk about this specifically-- how people assume causation when they see two correlated variables in a graph. We need to tread carefully with our graphs if we don't want to misinform our readers.--Urthogie 20:09, 28 December 2006 (UTC)

You are over-reading the APA report in this respect. The correlation between IQ and job status is mediated in part by education, but longitudinal research (some of it published just this year, and so beyond the time frame of the APA report) demonstrates that IQ and education each have independent inputs to occupational outcomes. The "training prospects" notion is better represented in military research than in comparisons of doctors, lawyers, chemists and ditch diggers. The U.S. military takes individuals (at age 17-18) from the top 90% of the IQ spectrum (noted in the training prospects column). Individuals below this level are too costly to train. Their extensive use of cognitive ability testing is a reflection of its utility in predicting trainability and job suitability even for an educationally homogenous (just HS graduation) population. Clearly the reason for mentioning training prospects is that it is one of the most important aspects of IQ differences for real-life outcomes, which has nothing to do with hereditarianism.

While it's within my abilities to dig out more examples from the mountains (seriously) of personnel psychology literature, there should be no reason to at this point. Your objections are moot, based on what appears to be a misreading of the literature and an undue emphasis on being contrary. --W. D. Hamilton 22:51, 28 December 2006 (UTC)

I'm not being contrarian, I'm being suspicious-- which makes sense based on the errors in the facts and POV that I have observed thus far. You say that military data is the best basis for the "training prospects", but if you actually look at the graph you'll see it doesn't just talk about who's ready for the military-- that's just one "prospect" out of many listed in the "training prospects" column. So where's the source for the non-military "training prospects"? Hell, where's the source for the military "training prospects", even? Find the sources, I'll promise not to edit in this regard until you get them, as long as you get them within a week.--Urthogie 23:50, 28 December 2006 (UTC)

Wonderlic 1992, the original citation, is the correct one for the non-military set of training prospects. The military training threshold is common knowledge, but is reviewed in a variety of sources, including the extensive Gottfredson (1997) [2]. I suggested you were being contrarian b/c the original Wonderlic citation was dismissed for reason that I think you now accept to be false. At this point, I don't see what the objection is. --W. D. Hamilton 16:52, 1 January 2007 (UTC)

You (seem to) have strong sources for the military claims. Perhaps we could include a seperate graph for how the military handles things-- as has been shown by above sources I've linked to and copied, the military training tests relate not just to intelligence. Perhaps there's some seperate source somewhere which demonstrates that the military tests somehow correlate to pure g, even when all other obvious relevant factors are addressed. I doubt it, but I'd like to see it, as it might warrant inclusion of the military stuff in the cohorts graph.

But more importantly, you don't have non-military sources for the "training prospects", and any sources you do find in this regard have already been a priori indirectly criticized by the APA report, which notes the limitations of correlating jobs with IQ (see above).

The reason I'm being so demanding on this data is that it's in graph form. It can easily mislead, as noted above.--Urthogie 01:21, 2 January 2007 (UTC)

I don't see what's insufficient about the Wonderlic citation, but I've supplemented it with the Gottfredson cite. --W. D. Hamilton 15:56, 2 January 2007 (UTC)

I'll look at that and get back to you.--Urthogie 19:32, 2 January 2007 (UTC)

Environmental explanations - cleanup from October 2006

I'll renew my concern that the Environmental explanations section is overly long and disorganized, and that the net effect is to obscure the most common and important arguments for this position. For example -- The 1st, 3rd and last two paragraphs present relatively uncommon (or unique) arguments, but their prominence suggests that they are of equal importance to the discussions of culture, toxic environments, and the Flynn effect. An attempt to discern which arguments are most important and to concentrate the text on those would improve this section. --W. D. Hamilton 17:34, 2 January 2007 (UTC)

Agreed. It might make sense to break down this and the genetic explanations sections into subsections for each of their main arguments.--Urthogie 19:34, 2 January 2007 (UTC)

WP:SS is in use on this article. I don't think the genetics section needs any more content -- explanation of the existing context or rebalancing of relative space is always a work in progress. The interested reader can get a explication of the genetic position from any of a number of prominent links. --W. D. Hamilton 19:43, 2 January 2007 (UTC)

It seems like we already make use of some subsections. Why would WP:SS not allow for more of this?--Urthogie 19:57, 2 January 2007 (UTC)

Considerations of size and "Levels of desired details". The existing subsections already push the meaning of SS to its limits. Ideally, a SS section would be about as concise as the lead block of an article is. I imagine that a 3rd level of headings would only invite further growth of the text, which would eliminate the macro/micro benefits of SS. --W. D. Hamilton 03:45, 3 January 2007 (UTC)

OK, I see what you mean. I guess we'll just try to organize the arguments through the lead paragraph.--Urthogie 21:10, 3 January 2007 (UTC)

schizophrenia

if we can't trim out the schizo example, then there's no hope for that section. why do we need an example about schizophrenia changing in immigrants? if environments are meaningfully different, then you can't generalize from WGH to GBH without extra data. this is knowable a priori. an example related to "race and intelligence" would be much better if an example is needed at all.

the intro should be able to tackle the basics, such as the universally agreed upon considerations. --W. D. Hamilton 07:44, 4 January 2007 (UTC)

Would you also be amenable to trimming out other specific examples not directly related to R&I, such as brain-size -> reaction time -> IQ? If we're removing indirect evidence on one side, it seems reasonable to remove it on the other side as well, and restrict examples only to direct measurements of race and intelligence. --JereKrischel 18:57, 4 January 2007 (UTC)

Nothing about the schizophrenia paper is directly or indirectly related to "race and intelligence"; it is not evidence for anything, but rather an example of a point for which there is actually no dispute. OTOH, a reading of papers by pro-hereditarians will reveal that they frequently discuss "brain-size -> reaction time -> IQ". My point is that the particular schizo example is unnecessary and irrelevant and adds nothing that a properly worded general description could not achieve. The claim it provides an example of is universally accepted and so could even go in the intro sub-section. --W. D. Hamilton 19:27, 4 January 2007 (UTC)

So, if there is indirect evidence for a link between R&I, we allow it, but if there is indirect evidence for the lack of a link between R&I, we don't allow it? AFAIK, Lieberman and others have made the point that heritability is not equivalent to genetic causality. And if it is "universally accepted" that heritability is dependent on environment, then shouldn't we mention that any claim that something is "heritable" is not particularly good evidence that something is caused by genetics? --JereKrischel 23:40, 6 January 2007 (UTC)

You're not getting my point. The sentence which the schizo example is an example for is not only perfectly relevant but universally (by all sides) agreed upon. It's the schizo example itself which has no relevance. The content of the example is totally uninformative about R&I. Better to have two sentences that carefully explain the point than one sentence to explain and one sentence to give an unrelated example. Moreover, we might as well put the claim in the intro paragraph because it is a notion upon which both arguments are based. --W. D. Hamilton 03:29, 7 January 2007 (UTC)

I'm not sure if you're explaining your point well either. Shouldn't we include an example which is "perfectly relevant" and "universally agreed upon"? On the other hand, if you'd like to start pruning the article, and removing "examples", I'd be more than happy to suggest the removal of other "examples" which seem misleading and irrelevant. --JereKrischel 08:17, 7 January 2007 (UTC)

The particular *example* isn't relevant *at all*. The *principle* that the example is supposed to exemplify (given in the sentence before it) is relevant. AFAIK, there's no reason at all for that example to be given -- no data was derived from that particular study that is any more relevant to R&I than any other study. It's not an example about race or intelligence; it's about immigrants and schizophrenia. --W. D. Hamilton 09:06, 7 January 2007 (UTC)

It is directly relevant to the idea of heritability, which is highly relevant to the idea of R&I. Given that the example exemplifies the principle mentioned before it, it seems particularly relevant and well founded. On the other hand, if you'd like to prune the article and remove examples which exemplify principles, I'd be more than happy to help. --JereKrischel 22:08, 7 January 2007 (UTC)

The section in question is in need of clean-up. If this sentence can't go, then I can't imagine you would allow anything else to go either. Is this section now set in stone, never to shrink to a reasonable size? Bigger simply isn't better. I resisted Urthogie's suggestion (above) that the genetics section include more detail for the same reason. A short and concentrated text is necessary and sufficient for readability in the explanations sections. If I find time, I'll try again to make improvements. --W. D. Hamilton 07:08, 9 January 2007 (UTC)

two questions for the editors

I said I'd give a week for each discussion past its end. The first removed claim has not been discussed since a week. Can I remove it?--Urthogie 20:00, 2 January 2007 (UTC)

I want to archive old discussions. How far back should I go?--Urthogie 20:00, 2 January 2007 (UTC)

gone for a while

I was just here to check up on this page for winter break. On my next vacation I'll continue discussion and apply what consensus is reached on. Sorry for the sudden leave, it's just that checking sources and whatnot isn't the most socialable activity during the school year. Till next time, --Urthogie 04:32, 5 January 2007 (UTC)

east asians in us and europe

here are the IQ score and citations for the measured average IQ of East Asians in North American and Europe since the 60s

103	Lesser et al., 1965
100	Coleman, 1966
96	Stewart et al., 1967
98	Werner et al., 1968
101	Peters & Ellis, 1970
103	Feldman, 1971
99	Flaughter, 1971
101	United States, 1971
98	Backman, 1972
101	Kline & Lee, 1972
101	Yee & La Forge, 1974
102	Winick et al., 1975
106	Winick et al., 1975
112	Winick et al., 1975
---
average = 101.5

101	Jensen & Inouye, 1980
115	Clark & Hanisee, 1982
107	Bagley et al.,1983
103	Vernon,1984
99	Sowell, 1986
107	Brandon et al., 1987
102	Pieke,1988
110	Frydman & Lynn, 1989
101	Flynn, 1991
110	Flynn, 1991
104	Jensen & Whang, 1994
103	Herrnstein & Murray, 1994
104	Lynn, 1996
109	Rushton, 1997
108	Stams et al., 2000
---
average = 105.5

those two averages are significantly different (p=.02) by T-test. the average of studies from East Asians in East Asia (since 1980) is 106. Why are the studies from ~30-40 years ago worth noting in in a paragraph-length exposition? I can't imagine you would have thought it a good idea to report the 60s/70s averages if it weren't already put there by someone else. --W. D. Hamilton 03:44, 7 January 2007 (UTC)

If you'd like to remove the older data which confounds the hierarchy asserted by folk such as Rushton, simply because of age, would you agree to remove the older crania/brain size/IQ data used to support the pro-hereditarian position? Can you also explain how the same citation has three different averages? (Winick et al., 1975) If in fact the data from ~30-40 years ago confounds the claims made by modern racialists, regarding a B-W-EA hierarchy, isn't that important, even if more recent data shows otherwise? It certainly seems to indicate a lack of consistency which would negate the possibility that race is a determinant of IQ, unless a race can change dramatically within a single generation. --JereKrischel 08:12, 7 January 2007 (UTC)

I simply can't believe that you think it's a *good idea* to include a paragraph about what the average IQ of East Asians was in the 1960s and 1970s when a full record of studies until the present finds that the numbers from the 60s and 70s do not accurately represent the present. (It's not the age but the apparent inaccuracy that is the problem. I also find it odd that you mention brain size measurements -- editing is not a game, no side "wins" -- biased editing only leads to lower quality articles.) AFAIK, there is no literature precedent to dedicate space to a debate about the magnitude of contemporary East Asian IQ. I also know of no studies which claims that East Asian IQ has increased since the 60s. (In contrast to to the BW gap.) The data we have at hand -- just the scores -- is compatible with that hypothesis or any of a dozen others that don't include a real change in scores. My personal guess would be that sample sizes from East Asians outside of East Asia have always been small and the low scores previously recorded are a result of poor sampling of a small population -- the scores of East Asians in East Asia are consistently around 106, where sampling isn't a problem. --W. D. Hamilton 09:04, 7 January 2007 (UTC)

I'm sorry, but what evidence do you have that the older studies are somehow inaccurate? Wouldn't the same logic apply to older crania/IQ studies of any sort? Should we avoid any studies that have small sample sizes (a common criticism of many of the pro-hereditarian claims)? It seems that a "personal guess" is the textbook definition of original research we should avoid. --JereKrischel 22:04, 7 January 2007 (UTC)

I mentioned my personal guess only to demonstrate that there is more than one possible explanation (you suggested another), not because I believe it should be part of the article. I don't particularly think the older studies are more technically inaccurate than the newer ones, only that the newer ones have the virtue of being more likely to accurately predict *current* values than the older ones (merely by virtue of temporal proximity; for example, a NY Times article published today mentioned that a lot of Asian immigration happened in the 70s/80s, which could be related to these data -- again merely mentioned only for the purpose of talk page discussion) -- I think this would be a non-controversial assumption. I'll try again to change the article when I have time, but there are a number of factors to balance, which the current text clearly fails to meet. --W. D. Hamilton 07:02, 9 January 2007 (UTC)

Do you then see this article as being "Current Race & Intelligence", rather than "Race & Intelligence"? I don't really see how it is appropriate to dismiss data because it doesn't predict any current group differences. Having different group differences historically than present is certainly a notable difficulty with the pro-hereditarian hypothesis (assuming that these arbitrary groups have any sort of essential nature beyond a few generations)...isn't that noteworthy? --JereKrischel 06:16, 10 January 2007 (UTC)

Having different group differences historically than present -- but we don't even have that. without a citation to back up that claim, we don't even have that conclusion to work with. all we have are older and newer review articles which give estimates of East Asian IQ with the newer ones giving higher estimates. there is no bigger "story" available to us than the existence of contrasting review article, with no contemporary review article arguing for a low number (or arguing for change or supporting the older estimates as true). so we can make no assumptions, other than the assumption that the latest data should be a more accurate reflection of the current situation.

drawing the "change" conclusion that you suggest would be inappropriate (for policy reasons per above) and hasty for scientific ones. a similar pattern of changing numbers exists for heritability estimates of IQ among whites. Early (date of publication) estimates were lower than later ones. It's not that the heritability increased over decades but who and what was measured improved. --W. D. Hamilton 08:07, 10 January 2007 (UTC)

Asserting that "who and what was measured improved" is simply one conclusion, don't you agree? --JereKrischel 09:50, 11 January 2007 (UTC)

so what i did was put all the studies i had available into a graph. i used the publication date of the study and the average IQ reported. i think this is as far as we can go on this topic w/o violating WP:NOR, but I also think it falls within the bounds of NOR because the data itself is all complied by Lynn and no attempt is made to draw any conclusions -- it's just the data. --W. D. Hamilton 04:19, 11 January 2007 (UTC)

Presenting selective data to imply a position can be POV pushing, as I'm sure you'll agree. Selectively censoring data presented that contradicts a point just isn't appropriate, even if you truly believe that "who and what was measured improved".

Certainly, a strong critique of much of Rushton's meta-analysis is regarding historical measurements - would you feel proper in censoring the data he claims as support for his position, since in contemporary measurements other data has appeared? --JereKrischel 09:50, 11 January 2007 (UTC)

How does what you wrote relate to the current situation? Nothing selective has happened. Rather, I've taken the least selective root possible -- presenting all the data in such a way that people can draw their own conclusions. I think that presenting all of the raw data is the only way to get around the dilemma of not being able to say that the scores have changed and yet the scores being obviously different. People can now see exactly what the scores were like in different regions at different times. --W. D. Hamilton 09:59, 11 January 2007 (UTC)

btw - when you argue for a position which I think is contra WP policy by appeal to a criticism of / threat to the partly-genetic view (particularly when it relates to Rushton) it only makes me lose confidence in your seriousness as an editor. writing good articles (which means abiding by WP policy) is my goal here, not the marginal advancement of one POV or another. (there are much better ways to do that than writing WP articles.) --W. D. Hamilton 09:59, 11 January 2007 (UTC)

If all we did was present raw data, perhaps I would agree with you - the meta-analyses of the data which are presented as fact are really one of the biggest problems with the article. My point is that you are selectively repressing data that undermines the position you believe in, and seem unwilling to apply the same high degree of skepticism towards data that bolsters your position. I'm simply asking for a consistent application.

Of course, in many cases, meta-analysis is all that is available, and there are critiques and counter-critiques galore. But I would ask you to be willing to apply any principle you may have which would eliminate mention of data contrary to the pro-hereditarian position to mention of data supportive of the pro-hereditarian position. --JereKrischel 10:20, 11 January 2007 (UTC)

I have tried in this thread to outline what I believe to be a dilemma -- which I believe I found a solution to. I believe it is a unique dilemma caused not by an over-abundance of "critiques and counter-critiques", but from an apparent lack of comment in the literature. One horn of the dilemma is that we cannot say even indirectly (by the composition of what we do say) things that are not in the literature. Choosing to mention Flynn's 1991 book in the same section that we say "East Asians score around 106" would be indirectly claiming something about East Asian scores that we don't have the permission to claim. Addressing that dilemma directly, rather than making what appears to me to be tangential claims, would be the best way to collaborate. --W. D. Hamilton 16:51, 11 January 2007 (UTC)

OR paragraph

Over the years, there has been variation in both the observed average IQ of groups, as well as the relative relationships between the average IQ of groups. Early 20th century measures typically found Blacks on the low end, and Whites on the high end. In more than a dozen studies from the 1960s and 1970s analyzed by Flynn (1991, 2002), the mean IQs of Japanese- and Chinese American children were always around 97 or 98; none was over 100. These studies did not include other Asian groups such as the Vietnamese, Cambodians, or Filipinos; who tend to under perform academically and on conventional psychometric tests (See Flynn, 1991). More recent measures find Blacks still on the low end, but put Asians on the high end.

I assume you must not have understood my point about a dilemma, because this paragraph clearly commits to OR. The first sentence is what you think the pattern of the scatter plot means, but it is OR to claim that is the meaning. --W. D. Hamilton 16:59, 11 January 2007 (UTC)

I tried to fix this, as long as ultramarine doesn't remove my fix. --W. D. Hamilton 18:58, 11 January 2007 (UTC)

Seems i didn't fix it the last time. AFAIK, my version doesn't juxtapose studies in a way to suggest any particular conclusions other than those of the constituent studies, which would violate NOR as the previous paragraph did. --W. D. Hamilton 19:17, 11 January 2007 (UTC)

No OR is happening here, this change in hierarchy is clearly cited by lieberman. It is presented here with no conclusion in a neutral manner. Lieberman actually is much more critical in his paper. --JereKrischel 19:21, 11 January 2007 (UTC)

[Ignorance] wasn't Lieberman talking about really really old studies of brain size? We're talking about IQ scores since the 60s. --W. D. Hamilton 19:23, 11 January 2007 (UTC)

http://www.ssc.uwo.ca/psychology/faculty/rushtonpdfs/Lieberman2001CA.pdf.

In the 19th century measurements of cranial capacity by Morton and others supported a “Caucasoid > Mongoloid > Negroid” hierarchy of intelligence. This continued through most of the 20th century but was challenged by a nonhierarchical view originating with Boas. Beginning in the 1980s Rushton correlated cranial and IQ measurements and presented a hierarchy with “Mongoloids” at the top.

They were talking about intelligence. --JereKrischel 19:27, 11 January 2007 (UTC)

Only in the sense of brain size = intelligence. This material is appropriate for the brain size section, but it should be obvious that trying to integrate it with Flynn (1991) is an OR synthesis. (Assuming Lieberman doesn't cite "changes" in East Asian or Black IQ as being part of his evidence.) In the end it's much easier to write when you don't violate NOR than when you do. The only intellectual demand on editors is to be reporters, not researchers. --W. D. Hamilton 19:32, 11 January 2007 (UTC)

JK, the changes you made are still a problem -- the OR still exists in trying to tie the BW IQ gap to the W-EA IQ gap in the same paragraph. The Lynn 1982 citation pre-dates Flynn and is necessarily talking about studies from the 60s/70s/early 80s. WP:NOR is quite clear about how we should be writing. --W. D. Hamilton 21:01, 11 January 2007 (UTC)

I disagree with your interpretation of WP:NOR. You seem to be attacking lieberman in a fairly legalistic fashion - he made very clear in his paper that the hierarchy of intelligence by "race" has done a peculiar thing by changing over the years, and to assert that "IQ" != "intelligence" really undermines the entire title of the article. If you'd like to change the article to "Race and IQ" instead, I'd be happy to oblige, and prune references to anything but "IQ" from the discussion. --JereKrischel 08:23, 14 January 2007 (UTC)

U.S./U.K gap

From the reference:

In my comments on the Black-White differential in Britain I was relying mainly on the following:
[1] Nicholas Mackintosh, IQ and Human Intelligence, OUP, 1998
[2] A. West, C. Mascie-Taylor, and N. Mackintosh, Cognitive and educational attainment in different ethnic groups, J. Biosocial Science, 1992, 24, 539-554
[3] N Mackintosh and C. Mascie-Taylor, The IQ question (1986), appendix to an official report on ethnic minority education in Britain.

The gist of the various studies, summarised in [1], is that early studies (1960s and early 70s) on Blacks in Britain tended to show average IQ around 85 (with Whites around 100): a similar differential to the US. However, the samples included very recent immigrants, and when these were stripped out, the gap was narrower. More recent studies (1980s onward) have shown a further narrowing of the gap. Notably, the Child Health and Education Study (1980) shows a difference of only about 5 points on non-verbal tests and 9 points on verbal tests.

Seems to be a valid reference. --JereKrischel 19:18, 11 January 2007 (UTC)

To MY blog!!! The underlying papers are valid, but the novel conclusions about UK/US are not. This data is actually likely best explained by the finding from Flynn that US kids also score higher than 85. It really should go, but I'm not going to remove it again today. --W. D. Hamilton 19:20, 11 January 2007 (UTC)

How is comparing UK/US numbers to each other a "novel conclusion"? Asserting a reason for the differential is a novel conclusion, perhaps, but simply stating that there are data showing a smaller B-W gap in the UK than the US isn't a conclusion at all, simply an observation. --JereKrischel 19:23, 11 January 2007 (UTC)

The US numbers were assumed to be 85. That assumption is probably false. But moreover, these papers don't compare US and UK numbers and conclude themselves that the UK numbers are smaller (AFAIK), else the poster would have just said so. The narrowing gap within the UK could be cited along side the Flynn data on narrowing (assuming it isn't part of his data already.) --W. D. Hamilton 19:27, 11 January 2007 (UTC)

So if we had a cite (which I'm sure we do) for the US B-W gap, you'd be okay with it? I'm not understanding your objection, the comparison of one B-W gap to another B-W gap is simple arithmetic, considering it a meta-analysis puts a great deal of the existing article in the same position. --JereKrischel 19:31, 11 January 2007 (UTC)

I'm saying that the conclusion that the UK and US BW gap are the same or not the same can only be made by a published paper. Assuming that kind of comparison doesn't exist, estimates of both gaps can be discussed at the same time providing they are comparable kinds of studies, so long as an implicit contrast (saying which is greater) isn't built into the discussion. --W. D. Hamilton 19:35, 11 January 2007 (UTC)

Strongly disagree A published paper is not required to assert that 25 is not equal to 15. Asking for a published paper regarding trivial math is like asking for a published paper regarding any placement of two numbers together in the article. --JereKrischel 19:38, 11 January 2007 (UTC)

If all you had to do was compare two numbers I would agree. But it would require at least a little statistics to say that the two numbers *are* (significantly) different, rather than trivially different. This is why we ourselves can't conclude that the 60s east asian scores are lower than the 00s asian scores. --W. D. Hamilton 19:42, 11 January 2007 (UTC)

The difference was not characterized as "significant" or "trivial", it was simply stated, neutrally and accurately, as "different". We ourselves can clearly conclude that the 60s east asians scores are lower than the 00s asian scores - we may not be able to conclude that these scores represent anything real, or have a certain cause, but nothing about statistics precludes the application of such trivial comparison. --JereKrischel 19:51, 11 January 2007 (UTC)

No as stated, it included 2 random variables (US IQ and UK IQ) and so statistics are needed to distinguish. You can say that the UK average is reported to be higher than 85, but you can't say it's higher than some other number which is a variable (not a constant) -- not without a study which draws that conclusion and can bear the burden of proof. A blog maybe as good as a news story, but it can't be as good as a research paper/book. --W. D. Hamilton 20:58, 11 January 2007 (UTC)

First of all, your application of logic here is misleading - the B-W gap has been measured in the U.S., and measured in the U.K.. Once measured, these values are no longer variable, but calculated. There is no need for complex statistics to report that these two calculated values are different.

Secondly, are you asserting that there is a different average IQ in the U.S. versus the U.K.? Are you suggesting that the average U.S. white IQ is different than the average U.K. white IQ? Do you have a citation for that? Could we make a note of this when talking about the issue, making it clear that even among "white" people, there are world-wide differences in IQ? --JereKrischel 08:28, 14 January 2007 (UTC)

The location where the text in question currently exists provided examples of the same group having different IQ scores in different places (credit to Lynn mostly). Comparison of UK and US in that paragraph would be an inappropriate contrast. --W. D. Hamilton 19:36, 11 January 2007 (UTC)

Violations of NPOV and dubious factual accuracy

POV descriptions

The article makes several misleading descriptions which gives the impression that one particular view, the pro-genetic side, is correct.Ultramarine 22:21, 11 January 2007 (UTC)

"Charles Murray has concluded that the U.S. Black-White gap on the SAT has increased in size to 1 SD". In same paragraph, other researchers findings are described as "argued".Ultramarine 22:21, 11 January 2007 (UTC)
"On average, a difference of approximately one standard deviation was observed in the US between the mean IQ score of Blacks and Whites." and " Whether these gaps are narrowing or not is debated." Very misleading, many researchers argue that they have narrowed, not that they may narrow as the text implies. Should be changed to "On average, earlier studies found a difference of approximately one standard deviation was observed in previous studies in the US between the mean IQ score of Blacks and Whites." and " Whether these gaps have narrowed or not is debated."Ultramarine 22:21, 11 January 2007 (UTC)

(1) Is there a dispute about Murray's conclusion? Is there a counter-"argument"? (2) Current studies continue to find a gap of 1 sd for adults. The 2nd sentence could read "There is disagreement about whether the results of IQ tests show a narrowing of the IQ gap, or if they do who has benefited." This is the sentence a few paragraph down where the subject is detailed, so it's all actually a bit redundant. --W. D. Hamilton 22:37, 11 January 2007 (UTC)

1. Again, the problem is the different language. NPOV requires equal treatment. Either all "argue" or all "have concluded". 2. Dickens and Flynn finds much less than 1 SD today.Ultramarine 22:49, 11 January 2007 (UTC)

(1) You're suggesting that every word must be identical, even when common sense language choices suggest otherwise? (2) You are misreading Dicken's and Flynn. Their work is well summarized in this article. You will find that they are discussing an effect on children and that they conclude that the adult BW gap is 1.1 sd. --W. D. Hamilton 22:52, 11 January 2007 (UTC)

Different words suggests different meanings. What is the problem with using my suggestions?Ultramarine 22:54, 11 January 2007 (UTC)

The only problem is that you argued that it was somehow demanded by policy. Argue is probably fine, but "claim" was over the top. --W. D. Hamilton 22:56, 11 January 2007 (UTC)

Yes, NPOV demands this. Regarding the gap, the text does not state that it applies to adults and it should certainly be mentioned that it is argued that some gaps have narrowed. Also, Nisbett argues that it has narrowed also for adults.Ultramarine 23:23, 11 January 2007 (UTC)

As a matter of principle, NPOV demands no such thing. Disputed claims may be claims or arguments for the sake of describing them as controversial, but non-disputed claims are clearly the conclusions of their authors.

Children are always a special case, and can be indicated as such -- for example in heritability. The intelligence of children is intertwined in developmental processes. The intelligence of adults is largely fixed until senility.

Claiming that only earlier studies find a BW gap of ~1sd is a vast over statement. Studies of adults still find a gap of ~1sd. --W. D. Hamilton 23:49, 11 January 2007 (UTC)

NPOV demands that all views should be given equal weight. "Have concluded" implies that this particular researcher is right, "argue" a more doubtful position. Again, Nisbett finds lower SD differences.Ultramarine 08:12, 12 January 2007 (UTC)

I don't see any difference in the use of the terms here. Both make clear that the article itself endorses neither position. By the way, "claims" is a word to avoid. --causa sui ^talk 17:01, 22 January 2007 (UTC)

Then there should be no problem with stating that Murray argues and Nisbett have concluded.Ultramarine 18:06, 22 January 2007 (UTC)

If there were a connotation difference with no actual difference between what's being described, then that would be unbalanced. But one set of words describe disputed broad conclusions (argue) and another describes a single data point that's not disputed (conclude). If no one disputes you, then are you arguing? It's a trivial point, but then again, it's a trivial point. It probably wouldn't have come up if you hadn't tried to change conclude to claim, which is weaker than argue.

Nisbett only looks at achievement tests -- hence children. His results looks just like Flynn's in that respect, only Flynn has high g loaded tests and more of them. Moreover, Nisbett is writing his opinions about the BW gap in a non-peer reviewed setting -- in a rebuttal to Rushton and Jensen. Thus, it doesn't have the rigor of the Flynn/Murray debate, where each is writing original research articles to make their points and are subject to the strictest standards. It's a bad idea to try to squeeze Nisbett's paper into the claim that there's a active dispute about whether adult BW gap is ~1sd. --W. D. Hamilton 18:17, 12 January 2007 (UTC)

This is part of the discussion regarding the stability of the gap. It is not the role of Wikipedia to decide who is right. If we can include Lynn's trade book, then we can include Nisbett's article. No double standard please.Ultramarine 18:27, 12 January 2007 (UTC)

Nisbett's conclusions are spelled-out where appropriate, no? We're talking about a summary sentence earlier in the text. You added "earlier" to qualify studies which find a BW gap as ~1sd. What should actually be qualified is "adults", not "earlier", as that's where the real agreement is. What is debated is ealier versus later among children. My point about Nisbett is that he has to be read in the context of the other studies. Nisbett probably didn't think to distinguish adults and children, but that's the point of Flynn's later work (and Murray's), and thus we should take Flynn's refinement of Nisbett's observation as part of an evolving body of literature. --W. D. Hamilton 19:26, 12 January 2007 (UTC)

How about the current text? Ultramarine 12:57, 13 January 2007 (UTC)

The footnote you restored is outdated as it doesn't include the newest material. As the intended content of the footnote is identical to the new section of the BW gap, there's no reason for a footnote. It's only 2x as hard to write neutral text if we have to have it written in 2 places. The details of the children/adults text don't precisely match what Flynn found. Flynn thinks that the gap varies continuously with age. --W. D. Hamilton 16:28, 13 January 2007 (UTC)

The text does not deny this. It should be explained that the gap may be narrower for those young which may translate eventually into differences for adults.Ultramarine 19:03, 13 January 2007 (UTC)

Except that Flynn right now says it doesn't translate. He thinks that the gap grows as individuals age. If Murray is correct, the gap got smaller amongst children c.a. 60s/70s and then fixed at the "new" normal developmental pattern. Neither says that the gap among children translates into a smaller gap among adults. --W. D. Hamilton 19:23, 13 January 2007 (UTC)

He says that the current gap widens for adults, not that it will also widen to the same degree for those currently children.

You can get his and Dickens' current thoughts by watching the debate. Flynn finds the fact that the gap changes with age to be an important result, but doesn't think they are any smaller among adults. He thinks this implicates environmental causes. I don't see anything in him predicting that the adult IQ gap will change; in fact he suggests that that will only happen with social/government intervention programs. --W. D. Hamilton 19:36, 13 January 2007 (UTC)

He is stating that there may be even more gains with changed environment. The current text is neutral, the change among young blacks that Dickens and Flynn see may or may not remain as they grew up.Ultramarine 16:38, 14 January 2007 (UTC)

Whether these gaps have narrowed or not, especially regarding children which may or may not later also cause a narrowing for adults, is, as noted below, debated.

The bolded part is the main problem. The relation between child and adult gaps is being described in a way that isn't supported by my reading/listening to the sources we have at hand. Flynn and Murry don't seem to go into this kind of debate - rightly b/c the data doesn't support any conclusions - and so the bold text is not justified. Saying that age is a factor in the debate -- a previous version -- would circumvent the specificity problem of the current text. --W. D. Hamilton 20:47, 14 January 2007 (UTC)

Information about the Pioneer Fund excluded or hidden

Exactly why the Pioneer Fund is criticzed of is hidden in footnotes while the arguments of defenders are presented in the main body.Ultramarine 22:21, 11 January 2007 (UTC)

To what does this refer? --W. D. Hamilton 22:38, 11 January 2007 (UTC)

The material previously in the footnotes.Ultramarine 22:49, 11 January 2007 (UTC)

I still can't tell what you mean? --W. D. Hamilton 22:52, 11 January 2007 (UTC)

If you see no problem with the current text, fine.Ultramarine 23:03, 11 January 2007 (UTC)

I just don't know what text is under dispute. --W. D. Hamilton 23:50, 11 January 2007 (UTC)

If you see no problem with current text (only difference is moving of material from two footnotes to main body), then fine.Ultramarine 08:09, 12 January 2007 (UTC)

i'll have to review the edit history. i can't remember that section well. --W. D. Hamilton 18:17, 12 January 2007 (UTC)

Keith Booker, a political figure, isn't an appropriate spokesman for scholarly opinion. His opinion that "this research is being done in the name of white supremacy" is being conflated with those of scholars like Tucker. The totality of the presentation suggests that it is a common opinion among scholars that hereditarians have neo-Nazi sympathies. Relatedly, there's an inappropriate blending of criticisms aimed at PF and those aimed at those funded by PF.

Secondarily, JK's formulation of titles regarding the "utility of race" should be reverted. They are part of a push on his part to claim that the "utility of race" is a central question in this field. This is largely due to his over emphasis of the opinions of certain anthropologists. --W. D. Hamilton 19:42, 12 January 2007 (UTC)

A third concern is the breakdown of distinction between the sections "Utility of research and racism" and "Accusations of bias". It doesn't that that much space to lay out what has and hasn't been said about "Accusations of bias". Charges that the research is about racism doesn't require two separate treatments. --20:10, 12 January 2007 (UTC)

Opinions regarding the goals of organizations and politics in general does not need to be published in peer-reviewed articles.Ultramarine 13:07, 13 January 2007 (UTC)

I'm talking about WP:NPOV not WP:RS. --W. D. Hamilton 16:29, 13 January 2007 (UTC)

If some significant argument is missing, add it. NPOV is not an argument for excluding some views.Ultramarine 19:06, 13 January 2007 (UTC)

NPOV requires a hierarchy of views (by, roughly, commonality) and a delineation/distinction/attribution of views (not a mixture), neither of which is being satisfied by the current text. The previous text said roughly x are the major critics of the pf and Keith Booker was left as example of the range of views. NPOV also doesn't require that some particular fact appear in any particular article, only somewhere in WP. not everything that's been said about PF is worth writing in this article. --W. D. Hamilton 19:09, 13 January 2007 (UTC)

What is wrong with him as an example of this view of the fund?Ultramarine 19:16, 13 January 2007 (UTC)

For example, he's not distinguished from Tucker. They are given equal weight. The previous text was an attempt to correct this implied equivalence. Everyone cites Tucker; no one cites Booker (obviously). That distinction makes Booker nearly irrelevant. --W. D. Hamilton 19:21, 13 January 2007 (UTC)

It is clearly stated who Keith Brooker is. If there is something missing regading who Tucker is, add it. The view of the National Association for the Advancement of Colored People is a significant view.Ultramarine 19:26, 13 January 2007 (UTC)

Booker is the president of the Delaware chapter of NAACP. He doesn't represent the group. Moreover, is it a common view among scholars that hereditarians are motivated by Nazi-beliefs? --W. D. Hamilton 19:33, 13 January 2007 (UTC)

Tucker certainly notes Nazists among those funded. Regardless, again, you do not need to a scholar to criticze an advocacy group, implications of research, or possibly causes for doing research.Ultramarine 19:39, 13 January 2007 (UTC)

Tucker certainly notes Nazists among those funded. Of course that's quite different than what's said by Booker. you do not need to a scholar - you don't need to be, but it does make a difference. the actual point i was making is the NPOV criteria of majority/minority/fringe views. which does booker represent? which does tucker represent? by juxtaposing the two, it is implied that both represent a common view, and moreover that both are canonical representatives of the views they espouse. NPOV is pretty clear on all this. --W. D. Hamilton 08:07, 14 January 2007 (UTC)

The text is not stating that these persons have the same view. The readers can form their own opinions, considering who they are are and what they state.Ultramarine 16:30, 14 January 2007 (UTC)

Juxtaposing them we create the impression that they have "equivalently" important views (not identical views). The previous text expliciting built the distinction of who are the major critics. Booker by an definition is not a major critic. --W. D. Hamilton 17:33, 14 January 2007 (UTC)

The previous text hide the criticisms while presenting praise in the main text. Juxtapositioning two authors do not mean equivalent, as it does not numerous other places in the article.Ultramarine 19:55, 14 January 2007 (UTC)

Equivalence of importance' or relevance or representatives is implied by ommision. You can't object to my objection merely by saying that the previous version was also not acceptable for some other reason. --W. D. Hamilton 20:14, 14 January 2007 (UTC)

I do not understand your argument. In the article in other places authors with different views are placed right next to one another without this meaning support.Ultramarine 20:36, 14 January 2007 (UTC)

My claim is that Booker (as the most extreme example) is not notable in himself as a critic of the research. AFAIK, this is the only thing he has every said about it -- a quote taken from a news story. At the very least, some kind of qualification about what views are taken as representing common versus uncommon views is required. The only context now is simple juxtaposition with quite notable figures. In other sections (hopefully), equivalently notable figures/views are present side by side, but non-notable figures/views aren't. --W. D. Hamilton 20:41, 14 January 2007 (UTC)

Something probably needs to be done with the "bias" section as per the discussion below. --W. D. Hamilton 20:42, 14 January 2007 (UTC)

How about removing that particular quote since the link is dead for me? Any other problem? Ultramarine 21:28, 14 January 2007 (UTC)

I'll review carefully if that change is made. --W. D. Hamilton 21:42, 14 January 2007 (UTC)

moved from above for attention - edits by ultramarine

"earlier studies" about BW gap - Flynn does not believe that the scores of adults have changed, only children.
"have narrowed" - flynn/murray debate appears to focus on whether the scores are currently narrowing among children or whether there was a one time narrowing among children
trade book - there is no fundamental difference between a trade book and a book published in an academic press. you can include published criticisms, but you can't eliminate things as central as lynn's book.
"conclude"/"claim" - there is no difference when each is cited and there is no dispute over the conclusion/claim, but claim does connote doubt where none appears justified.

--W. D. Hamilton 18:57, 11 January 2007 (UTC)

My answers is presented in context above.Ultramarine 23:02, 11 January 2007 (UTC)

boas error?

this concerns me about the uncritical reporting of boas' results: http://www.pubmedcentral.nih.gov/articlerender.fcgi?tool=pubmed&pubmedid=12374854

These results contradict Boas' original findings and demonstrate that they may no longer be used to support arguments of plasticity in cranial morphology.

these considerations, added to the age of boas' work, suggest that we must be very careful about what we say is actually established by boas. --W. D. Hamilton 19:05, 13 January 2007 (UTC)

That has been contradicted by later studies, see the Boas article.Ultramarine 19:07, 13 January 2007 (UTC)

At the very least, a debate exists and should be noted, but more specific arguments about the BW gap wrt this question have in fact been made. Here's an example:

Over 90 years after Boas’ (1912, 1940) classic study demonstrating changes in the descendants of immigrants, the extent to which cranial morphology reflects genetic variation remains controversial (Armelagos and Goodman, 1998; Sparks, 2001; Sparks and Jantz, 2002). The present results will not settle the issue, but may provide some additional insights. One could argue that significant changes in cranial morphology over the span of just 150 years are most likely due to plasticity. But, if this is true, they are confined to within-group change. The parallel course of secular change in Blacks and Whites, combined with their failure to converge upon a common morphology (see Figure 5), supports the hypothesis that genetic variation between the two groups is reflected in cranial morphology, despite exposure to a common environment. American Blacks and Whites undergo similar secular trends, but genetic differences between them are maintained.

http://web.missouri.edu/~wescottd/docs/Wescott%20&%20Jantz%202004.pdf --W. D. Hamilton 19:24, 13 January 2007 (UTC)

There seem to be no discussion regarding that there has been significant change in cranial size, similar to the Flynn effect. The implications are certainly discussed.Ultramarine 19:31, 13 January 2007 (UTC)

There seem to be no discussion where? I don't get your sentence? --W. D. Hamilton 19:34, 13 January 2007 (UTC)

All seem to agree that there has been an significant increase in cranial size. The implications and causes are certainly discussed, similarly to the Flynn effect.Ultramarine 19:36, 13 January 2007 (UTC)

If by "similarly to the Flynn effect" you mean that is has had no effect on the BW gap, that's really quite important, no? That more or less makes the results meaningless to this article, no? If not meaningless, then we need to spell out what people take to be the meaning if it's going to be presented. --W. D. Hamilton 19:38, 13 January 2007 (UTC)

Boas and later anthropologists certainly thought and find this important.Ultramarine 19:40, 13 January 2007 (UTC)

What's missing right now is (1) a modern citation to support the linking of Boas to the IQ debate (a cite from Boas would just be history), and (2) a proper accounting of the "within-group only" nature of the changes that Boas is argued to have found (Jantz 2004). --W. D. Hamilton 08:04, 14 January 2007 (UTC)

Again, see the Boas article here in Wikipedia for follow-up studies to the Jantz study. As noted in these articles, these findings remain valid and discussed also today.Ultramarine 16:40, 14 January 2007 (UTC)

AFAIK, you're talking about the wrong study. Should I assume that you've lost interest in this? --W. D. Hamilton 17:32, 14 January 2007 (UTC)

No, I have not. See Franz_Boas#Physical_Anthropology Ultramarine 19:53, 14 January 2007 (UTC)

The debate as outlined in the Boas article is fine and we should simply link to it. The article that seems to actually link Boas' findings to BW difference in brain size is the Jantz 2004 paper, which I quoted from just above. The findings are of no relationship, which certainly requires a change to the current text if Boas' findings are to be discussed at all. --20:05, 14 January 2007 (UTC)

Yes, you are discussing a differnt article by Jantz. But this article does not in any way contradict an increase in brain size. Some quotes:

"but changes in infant health and nutrition appear to be the most credible explanation for the craniofacial secular changes in American populations(van Wieringen, 1986)." " We agree with Angel (1982) that the relationship between cranial base height, nutrition, and health are real.""While there appears to be a strong relationship between cranial base height, nutrition, and health, changes in vault shape are probably not because cranial base growth is inhibited by the weight of the brain when health and nutrition are poor. Instead, as health and nutrition have improved, the brain has become larger and the vault has increased in height (Miller and Corsellis, 1977). To accommodate these modifications, there have likely been allometric changes in the direction of growth (Kouchi, 2000). Among American Blacks and Whites, increases in brain size and cranial vault height due to improved health and nutrition appear to be associated with a decreased growth rate in vault breadth and an increased growth rate in vault length.""Secular change in American crania is proximately related to a decrease in cranial base breadth and an increase in cranial capacity, and ultimately a reflection of improved infant growth due to better health and nutrition. The environment of 20th century Americans has no parallel in history. Activity levels are at an all-time low, and diet has improved to the point where overnutrition has surpassed undernutrition as our most serious malnutrition problem (Flegal et al., 1998). In addition, epidemic infectious diseases are now mostly controlled, and mortality is at an all-time low (Armstrong et al., 1999). It is not surprising that there is a biological response to this unparalleled environmental change."Ultramarine 20:45, 14 January 2007 (UTC)

this article does not in any way contradict an increase in brain size. Right, but it says that changes are confined to within-group change. The parallel course of secular change in Blacks and Whites, combined with their failure to converge upon a common morphology (see Figure 5), supports the hypothesis that genetic variation between the two groups is reflected in cranial morphology, despite exposure to a common environment. American Blacks and Whites undergo similar secular trends, but genetic differences between them are maintained. This seems to be crucial for this article, no? And so we should add this detail. But when this detail is added, is there anything verifiable that is still of interest about the brain size changes for this article? --W. D. Hamilton 20:52, 14 January 2007 (UTC)

He is not talking about cranial size here, which is the important thing. There might well be some genetic differences in cranial shapes.Ultramarine 20:54, 14 January 2007 (UTC)

Check the edits I made and tell me if I'm missing something here. --W. D. Hamilton 20:57, 14 January 2007 (UTC)

Again, Jantz is not talking about skull volume in that comment, but other forms of morphology. As he states elsewhere, "Among American Blacks and Whites, increases in brain size and cranial vault height due to improved health and nutrition"Ultramarine 21:07, 14 January 2007 (UTC)

Morphology --> size based on my first readings of the paper. The size is largely a product of the base plate morphology from my reading. I agree that Jantz appears to be agreeing that there has been increases in both Blacks and White. --W. D. Hamilton 21:13, 14 January 2007 (UTC)

No, he using various measurments, but not any size or volume measurment when making the statement you are citing.Ultramarine 21:18, 14 January 2007 (UTC)

He seems to be talking about all the morphological changes that underlie volume (e.g. vault height). His first PC is the one that separates blacks and whites. There's a lot of discussion of volume too, but note the text I wrote is only in relation to Boas. He directly cites his conclusions as having implications for Boas'. Is there something about this that I'm misunderstanding? Maybe you know/see something I don't. --W. D. Hamilton 21:35, 14 January 2007 (UTC)

That is your own original reserach, he using various measurments, but not any size or volume measurment when making the statement you are citing. In contrast, my quote is when he talks about size.Ultramarine 21:42, 14 January 2007 (UTC)

I'm willing to accept the possibility that I'm somehow confused about the entire issue, but the text seems pretty clear to me. Do you know of any additional articles that discussion BW differences in volume/morphology etc? --W. D. Hamilton 21:46, 14 January 2007 (UTC)

~~This is in the wrong sectin, please move.~~ No other articles really needed, the article cites numerous other articles. Again, when he talks about size, he notes an increase both for blacks and white probably due to nutrition.Ultramarine 21:52, 14 January 2007 (UTC)

when he talks about size, he notes an increase both for blacks and white probably due to nutrition - right, and that the persistence of difference must be due to genetics. what i'm asking is about publications that discuss BW differences in the context of longitudinal changes as Jantz 2004 does. is the only reason to cite boas that brain size varies with nutrition? if so, then the idea relative-morphology is probably maintained by genetic differences is important too. --W. D. Hamilton 21:58, 14 January 2007 (UTC)

He makes no such statemet regarding size and does not state that different morophology must lead to differences in brain size. Howeve, I think we could mention both the increase in brain size but also that other cranial measures are not converging. Obviously the finding that brain size has increased is very interesting for this field.Ultramarine 22:49, 14 January 2007 (UTC)

The text is improving, but "other" measurements is probably inaccurate. The PC analysis that is the basis for the BW conclusions isn't in terms of raw measurements, but rather a combination of multiple measurements. The actual factor loadings appear to be in another paper, so I'm not sure what those measurements are. But based on the text, they measure what changed about skulls over time and what is different about black and white skulls as groups. While volume is different in both cases (time and race), it shows that race differences are not the kind of differences that occur with time. AFAIK, this analysis is comprehensive of all considerations that go into a consideration of size -- it superceeds size. If I'm correct about this - by that I mean if I'm reading this in the way that it was intended to be understood - then that's quite a bit more than merely showing that "other" measurements didn't change between blacks and white. When I say morphology --> size, it appears to me that that is an implicit assumption of the discussion in the paper. This is why I asked for other papers on the same topic. Maybe they would clear up the issue. --W. D. Hamilton 00:21, 15 January 2007 (UTC)

titles of bias subsections

the titles used by JK in the bias section come from his redefinition of the hereditarianism wrt group differences debate in terms of the "utility of race". this is a result of his promotion of the views of (mostly cultural) anthropologists, especially lieberman, that race is the most important part of the R&I debate. most broadly, the race issue has always been secondary to the intelligence issue, and as our understanding of race has changed, the relevance of anthropologists versus geneticists to the issue has lessened. more specifically, the content of the section has absolutely nothing to do with the debate about race, but it is largely and clearly about "political correctness" (this is the term used and understood by all sides and by the reader) and the "pioneer fund". i assume that ultramarine agrees with the more specific claim at least. --W. D. Hamilton 08:02, 14 January 2007 (UTC)

I'm sorry, but I don't think it is appropriate to assert that the "race issue has always been secondary to the intelligence issue". If we were simply talking about intelligence and IQ, you wouldn't have nearly the controversy, as I'm sure you'd agree. "Political correctness" is a loaded term, and limiting a section title to "Pioneer Fund" seems misleading. --JereKrischel 08:19, 14 January 2007 (UTC)

I said "broadly" and then pointed out the "specific" point. There's nothing loaded about "political correctness" that isn't loaded about "racism" but those are the terms most commonly used. The PF section is just about the PF right now. It's certainly not about the "utility of race" which is quite specific (not just "race" but it's "utility"). --W. D. Hamilton 17:29, 14 January 2007 (UTC)

Would you agree to change it to "Anti-racist", instead of "Political correctness", if you find "racism" a neutral term, perhaps we could use that instead. --JereKrischel 19:18, 14 January 2007 (UTC)

No, I wouldn't find that appropriate. Anymore than I would describe pro-life as anti-abortion or pro-choice as anti-life. Each side deserves to have their terms used as self-described so long as it isn't terribly controversial. I see no objections in the literature to the use of the term political correctness in the context of human variation. The particular choice of "anti-racist" would suggest that hereditarians deem themselves to be racists. --W. D. Hamilton 20:09, 14 January 2007 (UTC)

The particular choice of "political correctness" would suggest that people who hold the anti-racist view deem themselves to simply be irrational left-wing reactionaries. I think we need to come up with something besides the terms bandied about by the various sides - there are clearly people on both sides of the argument who work diligently to use loaded and POV terms to add to their persuasiveness. --JereKrischel 10:13, 16 January 2007 (UTC)

the more sensible alternative would simply be to have no subsections, as this is (was, or could be) a very simple section. it has grown into a something quite larger b/c of the pioneer fund content. some redistribution and rethinking is probably a better use of time. --W. D. Hamilton

It seems clear that if we want to talk about "Accusations of bias", we have two types - the accusation that people dispute and denigrate the research because they're "politically correct", and the accusation that people blindly support and believe in the research because they're "racist". Obscuring that there are accusations of bias going both ways isn't appropriate. --JereKrischel 10:25, 15 January 2007 (UTC)

I've reconstructed most the material from the ancient "public controversy" section. There is no good reason for trying to reconstruct "accusations of bias" now that it is clear that there are multiple aspects of the "public controversy" which reflect to some extent an "accusation of bias". Deleting the well cited racism section is totally unacceptable. Trying to make the public debate about the "utility of race" is unjustifiable given that the vast majority of the review papers and books written on this topic do not take that stance. A full restructuring of the public controversy section seems appropriate for the sake of summary style, but there's no reason to favor the previous groupings, where were obviously created too soon. --W. D. Hamilton 17:33, 15 January 2007 (UTC)

Your "reconstruction" is clearly POV pushing. You argue in an editorial style, quoting weak strawmen for the anti-racialist position. You draw extensive quotations from pro-hereditarians, fail to edit them for brevity, and then present their rhetoric without any balance at all. If you would like to engage in a reconstruction, I would suggest that you first start by adding critiques against the people who support the racialist position - you may do a better job of remaining in an NPOV tone if you're writing for a position you don't hold personally.

Also, please provide some sort of citation to back up your claim that "the vast majority of the review papers and books written on this topic do not take that stance". It seems that the vast majority of critiques expressed both by anthropologists and even geneticists, seem to be most focused on the utility of race as a proxy for genetics. --JereKrischel 10:04, 16 January 2007 (UTC)

You have to start somewhere. I copied material here and added a clean-up tag to note that it was in disarray. However, you cannot simply remove well cited material. NPOV frowns on removal for all but the worst text. I assume this text doesn't meet that very high threshold for removal. As I pointed out -- the copying was done so that we can reassess what material we're supposed to be summarizing in these sections. I didn't have to do extra writing -- as implied by your suggestion that I have to add published rebuttals for the material I added -- except for the papers in press I added.

Race -- you are confusing discussions of race in other contexts than the discussion of intelligence and discussions of race in the context of intelligence. As a background material, the topic of race can be summarized in itself. As a structure for this article it is -- sorry to have to say it given the recent discussion, but it's true -- NOR to synthesize material that is not more directly linked (WP:SYNT). It is sufficient for thiings to be linked if a published report connects them. It's also sufficient as far as I'm concerned for there to be a good common sense connection. However, we have positive counter examples against a common sense argument. The various reviews of race differences in intelligence fail to characterize the public dispute in terms of the "utility of race". In the context of what the "utilty of race" dispute is about, it would be very odd to try to draw lines that way. The real debate is about whether there is a genetic cause to group differences. Race could any level of "utility" in the geneticists/anthropologists debate, and that would still be the question of interest (unless you subscribe to the view that there are no genetically caused phenotype differences -- not even skin color or other "racial" traits -- which is a fringe view). --W. D. Hamilton 18:05, 16 January 2007 (UTC)

You said, The various reviews of race differences in intelligence fail to characterize the public dispute in terms of the "utility of race". This is clearly a false statement, given even just the one example of Lieberman. The real debate about "whether there is a genetic cause to group differences" cannot fail to include a strict scrutiny of the use of race as a proxy for genetics, and this scrutiny is applied by a number of geneticists, psychologists, and anthropologists working in the field. Although this framing of the argument may be threatening to your personal position, and you may find it as "weaselly" because it seems like an indirect and "sneaky" attack, it is nonetheless well presented in the literature, and is at the core of the difference. --JereKrischel 02:28, 17 January 2007 (UTC)

(1) You are confusing the idea of discussing a topic with the idea of using a topic as the general framework for a discussion. The "utility of race" is sometimes a topic of discussion, but "whether there is a genetic cause to group differences" is the framework in which the dispute is discussed in the vast majority of publications. Citing the example of Lieberman is only evidence for the former, which I obviously am not denying. (2) Your descriptions re: the "utility of race" demonstrate a lack of familiarity with the details of the discussion. I don't blame you for this because writing in this area is particularly cryptic. Suffice it to say, your suggestions fall short of the mark. The intersections of "race and intelligence" and "utility of race" are specific and narrow. They do not form a backbone for the topic. (3) The best possible solution and direction for all NPOV disputes is to seek narrowly-tailored and rhetorically clever solutions. Trying to get "utility of race" into the titles of sections that are predominately about "political correctness" and the "pioneer fund" is not a solution. My use of those terms in that last sentence is not meant to suggest that those are the only acceptable terms. What isn't acceptable is to make the "utility of race" a central axis of the article or related articles -- it fails many tests and is unwise. Look for narrow and focused titles and I believe this dispute can be easily dissolved. --W. D. Hamilton 05:37, 17 January 2007 (UTC)

(1)"whether there is a genetic cause to group differences" is the framework in which the dispute is discussed in the vast majority of publications - Please provide evidence for this. It seems that it is true if one limits oneself to pro-hereditarian publications, but does not seem very true in general.

(2)The intersections of "race and intelligence" and "utility of race" are specific and narrow. I believe you don't have so much of an intersection as you have a foundation and a structure. "race and intelligence" requires a basic foundation of the "utility of race". To attempt to avoid the strongest critique against the various R&I conclusions made by pro-hereditarians seems inherently POV pushing.

(3) I agree that we need narrow and focused titles, but I don't see how "political correctness" is more focused than "those who dispute the utility of race". The "utility of race" is an obvious axis that avoids the loaded terms used by both sides, and provides a neutral POV. One can believe that race is a good proxy for genetics, and not be a racist, just as one can believe that race is not a good proxy for genetics, and not be left-wing PC fascist. Much of the victimization material produced by the pro-hereditarian side seems to be a fairly transparent ploy for sympathy, and may not be very appropriate for the article -> certainly not in the quantities as current. We should probably endeavor to prune these sections into much shorter summaries, with an equal number of examples and an equal weight of article space for both sides. One of your particular weaknesses on this topic seems to be the ability to balance the article appropriately - although your contributions are certainly helpful, they would be more helpful if you balanced your efforts to include equal weight to both sides of the issue. I know this is particularly difficult if you don't believe there is any rational basis for opposition to your POV, but it is possible. --JereKrischel 08:26, 17 January 2007 (UTC)

(1) the big 3 -- APA, WSJ and S&R

(2) I believe you don't have so much of an intersection as you have a foundation and a structure. - this is a POV whose popularity is hard to judge, but it fails to find any but the tiniest expression in any of the big 3 from (1). It's a position that makes little sense to me, not because I endorse some particular view on the "utility of race" (a multifaceted topic), but because they are simply orthogonal for the most part. That is, the extent to which race differences in a phenotype are caused by genetic differences is mostly orthogonal to the question of the size and stability and biological underpinnings of those phenotypic differences. Consider race differences in heart disease. The "utility of race" has little to do with the size of those differences, the stability over time and place, or the biological underpinnings. It does tells us a little about the a priori probabilty of a genetic differences, but does so very indirectly so as to be almost uninformative and thus any combination of genes and environment is possible (mountain and risch, 2004). Gould has said a few things along these lines (an interview with PBS), but again they only reflect on the question of genetic causation.

(3) Because the "utility of race" isn't mentioned in the section, isn't mentioned in any of the papers cited in the section, and I would imagine there are very few papers if any that mention both the "utility of race" and the "pioneer fund" in the same paper. so there's nothing about about the "utilty of race" being the axis to divide up these sections

(3b) "We should probably endeavor to prune these sections into much shorter summaries" - I believe summary style is our goal, but it's hard to summarize if you don't know what you're summarizing

"with an equal number of examples and an equal weight of article space for both sides." - Maybe justified for other reasons, but that's not what NPOV means. We don't "give equal time" to unequal positions. I stand to be corrected, but few people have been persecuted or called racists for suggesting positions we might call anti-hereditarian. We should make the article represent the majority and minority views that are commonly found in the literature -- the definition of NPOV.

(3c) know this is particularly difficult if you don't believe there is any rational basis for opposition to your POV, but it is possible. -- don't be silly. i think there are serious and highly respectable figures who believe both sides of this debate (and non-respectable ones as well). my bias is my knowledge and training in genetics and a view of the definition of "racism" that is similar to the one presented by Pinker. let the best science win. --W. D. Hamilton 02:38, 18 January 2007 (UTC)

p.s. This is not for the article, but as an expansion of 3C - a large number of famous and less famous geneticists privately believe the partly-genetic hypothesis, describing it as "the world's best-kept open secret". they also discuss amongst themselves and usually tend to conclude that it would be better on the whole for society to remain misinformed or even purposely mislead. sometimes this discussion is done in the open, as documented in the "utility" section. i don't know how i feel about their conclusion, but given that most people will dismiss this article as racist without reading it (and some leave us notes to that extent), i don't fear any harm to society will come from our efforts. --W. D. Hamilton 02:56, 18 January 2007 (UTC)

brain size paragraph that cites lieberman

Also, I didn't cite Deary, the review article in Heredity did. Not sure if it is our place to argue the point with the reviewer of IQatWoN - definitely seems OR, don't you agree? --JereKrischel 03:15, 19 January 2007 (UTC)

The entire paragraph appears to not be supported by citations. There's nothing in lieberman about reaction time that i can find. The book review is criticizing the use of reaction time measures in international comparisons. I don't see how this can be said to directly relate to the reports of reaction time differences by race in the primary research literature. Moreover, we can't be complicit in misquoting a source. It is within our power to leave out a misquoting if it's obvious. In this case it is. The only NPOV way around it is to say that the opinions are those of the reviewing author, but since he's talking about reaction time in a different context, that won't work. There may be criticisms of reaction time, but we'd need to look outside of book reviews of IQatWoN. Papers that cite the original studies could have criticisms, but I've never seen them mentioned in any reviews. --W. D. Hamilton 03:21, 19 January 2007 (UTC)

Finally, weak and confusing correlations between IQ and reaction times, which L&V naively take as an index of 'the efficiency of the brain' (p 66), are not 'theoretically tractable - This clearly is the critique outlined in the paragraph. And the source is not being misquoted at all. You may interpret the citation as a misquote, but I certainly don't agree with you. --JereKrischel 03:30, 19 January 2007 (UTC)

The argument is being used in a different context there. The purpose of RT data used by Rushton and Jensen, contrary to the claim which is currently attributed to "some" scientists and cited to Leiberman, is to show that the IQ differences bleed through to the underlying biological correlates of g. Criticisms of that point would be on target. Claiming that RT is not "theoretically tractable" isn't a criticism aimed at this point. At the very least, we need to have concordance between criticisms and the underlying claim being criticized. --W. D. Hamilton 03:39, 19 January 2007 (UTC)

Your interpretation of whether or not Liberman's assertions are contrary or not to the intentions of Rushton/Jensen definitely fall within OR. Lieberman makes the critique, and we report it - arguing Rushton's case isn't our job. Also, the IQatWoN review in Heredity clearly makes hay out of the "weak and confusing correlations". The underlying claim is that race -> brain size -> reaction time -> IQ. The critique of that claim occurs on both the level of "race -> brain size", as well as "reaction time -> IQ". There is concordance. --JereKrischel 03:49, 19 January 2007 (UTC)

I noted that what you claimed are Leiberman's views are contrary to what R&J have written. The actual problem is that they are not even Leiberman's views -- as far as I can tell -- you'll need to point me to the quotation from Leiberman. You appear to be oversimplifying and than connecting together topics that are actually distinct. For example, race -> brain size -> reaction time -> IQ is completely contrived, but also false. Although not a problem right now, what you call a criticism of "race -> brain size" is actually just an instance of the blanket criticism of "race" that happens to include brain size, but could include any phenotype including skin color (not the lessen the importance of the argument, or the fact that it has been said about brain size). The criticisms of "reaction time -> IQ" are limited to saying that they don't explain anything causally, which AFAIK is not a view being presented in the article. When I say claim and criticism need to be concordant, this is part of what I'm getting at. --W. D. Hamilton 06:33, 19 January 2007 (UTC)

For example, race -> brain size -> reaction time -> IQ is completely contrived, but also false. If it is completely contrived, why do we have a section dedicated to illustrating that connection? Am I imagining things, or are you agreeing with me? It seems that you're asking for a specificity in some cases that isn't required -> if the criticism of race->brain size is part of a blanket criticism that may include other traits, what problem is there? Are we to say that blanket criticisms that apply to more than one claim cannot be applied? --JereKrischel 21:09, 19 January 2007 (UTC)

No, I'm not agreeing with you. I'm saying that your misunderstanding the subject, and thus (seemingly) inadvertently writing false claims. For example, there's no discussion of reaction time by Leiberman. Read my comment in that light. --W. D. Hamilton 07:01, 20 January 2007 (UTC)

In the brain size section, the last paragraph is mischaracterizing Leiberman, but the other paragraph are pretty accurate. Do you see this now? --W. D. Hamilton 07:10, 20 January 2007 (UTC)

In other words, I think you've simply made several mistakes. --W. D. Hamilton 17:28, 20 January 2007 (UTC)

Upon further inspection, I think we have a larger issue here -> the citations given for reaction time studies simply do not address race. Including those references seems patently OR. --JereKrischel 08:21, 21 January 2007 (UTC)

And studies that look at the correlation between brain size and IQ do not at the same time look at race. However, the existence of such correlations is necessary background material to explain why race differences in these variables is relevant. This link is supported by every paper that looks at about race and brain size or race and reaction time in the context of intelligence. Thus, this isn't SYNTHESIS and can't be OR. --W. D. Hamilton 08:26, 21 January 2007 (UTC)

The assertions that such background material are supportive to a given point of view is speculative and OR at its worst. It would be just as appropriate, by your logic, to sprinkle in, at every mention of the word "race", the strong disagreements people have with the very utility of the concept as a proxy for genetics -> it is definitely necessary background material to explain why the use of those categories is a mistake. This concept is supported in hundreds of papers critical of the genetic hypothesis. Would you favor including that kind of SYNTHESIS? --JereKrischel 08:36, 21 January 2007 (UTC)

I don't understand why you don't understand, but the example you gave is completely different than the disputed situation. In one case, a piece of background data (correlation with IQ) is given. This background is usually if not always given in papers when the race difference is described. It presents the motivation for work being described, establishing the relevance of the variable to intelligence, but has no further part to play. It is appropriate in one and only one position in the article, and is not a disputed finding in itself. In the example you gave, you are suggesting that your favored view about race be interjected at points in the article where it would wrongly imply that a debate has been carried out in the literature about that particular topics in the article wrt the utility of race. The points in the article would be those that you chose on the basis of your own opinion. The view you would present is itself so controversial and complex that it requires a >100k WP article to describe it (see race). --- The correlations are background, essentially introducing the definition of the variables of interest wrt intelligence, and their inclusion is supported by a body of literature. Your suggestion OTOH is truly SYNTHESIS. --W. D. Hamilton 08:51, 21 January 2007 (UTC)

In the example I gave, I'm simply giving "background data" -> information which repudiates the use of "race" as a proxy for genetics. Although I may not dispute the findings of some scientific study that asserts a relationship between reaction time and IQ, it is inappropriate to put that "background data" in without also giving some mention that the study may not be supportive of the genetic hypothesis because it did not address BGH. I think you cited earlier, It introduces an analysis or synthesis of established facts, ideas, opinions, or arguments in a way that builds a particular case favored by the editor, without attributing that analysis or synthesis to a reputable source;. If your example is not a violation, then why wouldn't the same be true for citing lieberman's critique of race as a proxy for genetics, every time you try to add a study which has made no claims on race? --JereKrischel 09:02, 21 January 2007 (UTC)

Again, I don't understand why you don't understand, but the example you gave is completely different than the disputed situation. If I repeated myself? SNYTHESIS means putting together two things that aren't previously put together. Citing Leiberman as a critic of studies that link reaction time to race is SNYTHESIS. Citing the correlation between reaction time and IQ is supported by the fact that the studies which look at reaction time differences by race cite that correlation as their motivation. It would be SNYTHESIS to cite essentially any other finding except that there is correlation between reaction time and IQ (anything that hasn't been directly linked to the topic previously). --W. D. Hamilton 09:08, 21 January 2007 (UTC)

Citing Lieberman as a critic of the use of race as a proxy is just as relevant "background information" as citing a reaction time to IQ study that didn't look at the concept of race. The fact that studies of reaction time and race cite reaction time to IQ studies is just as true as the fact that critiques of reaction time and race studies cite the cluster vs. cline data. You put yourself in an inherent POV pushing position when you assert that the only "background information" that can be inserted is that which supports your POV. How about this, can you give me an example of citing "background information" which would not support your POV? --JereKrischel 09:28, 21 January 2007 (UTC)

But Leiberman doesn't discuss reaction time. But the the brain size studies really are motivated by the findings of a correlation between reaction time and IQ. You see? The particular details of this situation make the one not just appropriate but necessary and the other not appropriate. Based on your reasoning, we need to delete everything in the intro block about IQ and race because it isn't directly about race and intelligence. Secondarily, on practical grounds, people reading the article need to understand what reaction time measures have to do with intelligence because writing about reaction time doesn't make sense without it. In contrast, they don't need the race/not-race debate brought up at every mention of the word race. --W. D. Hamilton 10:22, 21 January 2007 (UTC)

The motivation of the brain size studies seems just as relevant when dealing with the utility of race as a proxy for genetics. Simply saying that there is some background motivation doesn't address my issue of your POV pushing. Please, give me an example of any sort of inclusion that would be appropriate, but is against your POV, or unsupportive of the genetic hypothesis research you've cited? I believe you're simply trying to frame the question in such a way that it is impossible to include any sort of debunking, but permissible to include supportive yet indirectly related evidence. --JereKrischel 10:35, 22 January 2007 (UTC)

lieberman's critique of race as a proxy for genetics - somewhat of a tangent, but Leiberman's claims have a very specific nature and it's a mistake to try to appropriate them in new situation. it would be likewise a mistake to present them as the authoritative view on the topic of race, there are many other more influential voices. thirdly, it's a mistake to characterize lieberman's view as being a "critique of race as a proxy for genetics". it's actually a critique of using race (i.e. genetic clusters) instead of clines. --W. D. Hamilton 09:13, 21 January 2007 (UTC)

We're saying the same thing. If you want to say race==genetic clusters, and as such is a good proxy for the genetic traits related to intelligence, you'd still be making the same argument. Trying to assert that "clusters versus clines" is any different than "race is a poor proxy for genetics" is a stretch. --JereKrischel 09:28, 21 January 2007 (UTC)

"clusters versus clines" == "race is a good/poor representation of human biodiversity (genetic and/or phenotypic)". the debate is old, complex, on-going, and by many accounts totally mired in politics. --W. D. Hamilton 10:22, 21 January 2007 (UTC)

i've changed the paragraph to attribute the claims to lieberman, and specify that it is rushton that he is describing. i don't believe the lieberman paper actually supports the text in the article, but it will take another re-reading to be certain. i think lieberman is criticizing rushton for claiming that systematic variation by race in a number of traits (beyond just brain size) supports a genetic explanation of group differences, but i see nothing in lieberman related to the magnitude of group differences in IQ (which is how the current text reads). quotes that you think support the text would be helpful on the talk page. --W. D. Hamilton 11:41, 21 January 2007 (UTC)

Lieberman's paper includes views from more than Lieberman (Rushton has a section in there too). I've reworded the paragraph a bit more. --JereKrischel 10:35, 22 January 2007 (UTC)

rdi reviews

here are actual reviews of RDI with specific attention to the reliability of the results (not representativeness, which isn't important for RDI) --W. D. Hamilton 07:05, 19 January 2007 (UTC)

Lohein: http://dx.doi.org/10.1016/j.intell.2006.05.001

Are the numbers accurate? I checked a sample of 40 of the 615 rows in the IQ tables against their sources—selected mostly on the basis of being readily accessible to me. Result: 14 of the 40, about 1 in 3, showed discrepancies, although mostly minor ones. For example, there were 9 with discrepancies in Ns, such as using an overall N from the study instead of the actual N on which the particular IQ was based. In a similar number of cases, the ages or age ranges were a bit off. The IQs were harder to check, since often I did not have the original norm tables used, or there was some uncertainty in applying Flynn corrections, but when I could, my estimates and the values in the tables were typically within a couple of IQ points. I only came across one large discrepancy—an IQ 14 points too high (in Table 12.1, row 20; due, according to an e-mail from Lynn, to a clerical error in adding instead of subtracting a Flynn correction). The citations and references were, on the whole, accurate. In short: Yes, the general trends in the tables are probably dependable, if the assumptions regarding Flynn effects, etc., are correct, but it is prudent (as always) to check with original sources before quoting particular results.

note: table 12.1 lists studies of native americans --W. D. Hamilton 07:05, 19 January 2007 (UTC)

Mackintosh: http://dx.doi.org/10.1016/j.intell.2006.08.001

Even if we wanted the data, can we rely on Lynn to have given an accurate account of them? I do not pretend to have read the originals of more than a handful of the papers and books cited by Lynn, but it just so happens that I wrote two of them myself, and Lynn has simply got their data wrong. Table 6.2 gives the results of 13 estimates of the IQ scores of south Asians in Britain and Australia. The median score is said to be 89, and similar for Indians and Pakistanis (as is to be expected since they are the same racial group). In fact, three British studies have given the same IQ tests to Indian and Pakistani children, and in all three, Indian children have outscored the Pakistanis by 4–6 IQ points. Mackintosh and Mascie-Taylor (1986) reported IQ scores of 97 and 93 for 10-year-old children of Indian and Pakistani origin respectively, but Lynn incorrectly attributes both scores to Pakistanis, one said to be for children resident in Britain more than 4 years, the other for more recent immigrants. West, Mascie-Taylor, and Mackintosh (1992) reported IQ scores of 91 and 85 for Indian and Pakistani children, but in Lynn's table these have mysteriously turned into scores of 87 and 88. The errors may not be particularly important, and I do not know how typical they are. But they do not increase my confidence in Lynn's scholarship.

note: the version of the data table i received seems to have these corrections already. --W. D. Hamilton 07:05, 19 January 2007 (UTC)

If you've received some special data table, and compiled a graph from it yourself, surely you agree that's OR. --JereKrischel 21:10, 19 January 2007 (UTC)

A friend copied all the data from the book, he or I must have fixed the mackintosh numbers back in when this was first online. you can't be OR by having copied data from a book. your attempts to find OR are a waste of our time. --W. D. Hamilton 07:05, 20 January 2007 (UTC)

I'm sorry, but "a friend" hardly qualifies as a reliable source, don't you agree? Neither is it appropriate for you to correct the numbers -> that is definitely OR. --JereKrischel 07:55, 21 January 2007 (UTC)

Also, AFAIK, the word "reliable" only occurs in the R&I article here once...I don't think we have an argument about "reliability" that is relevant here. --JereKrischel 21:16, 19 January 2007 (UTC)

we don't have an argument about representativeness at all. by reliability i mean accuracy (which is actually a mix of reliability and validity). --W. D. Hamilton 07:05, 20 January 2007 (UTC)

The critique specifically address the representativeness of the data, not the reliability. I'm unsure why you're trying to bring up reliability at all, we don't have an argument about that. --JereKrischel 07:55, 21 January 2007 (UTC)

There are no criticisms of RDI wrt representativeness. It would make little sense to make such a criticism because of the very different nature of the work. Criticisms of representativeness aimed at IQatWoN do not naturally reflect on RDI as well. --W. D. Hamilton 08:22, 21 January 2007 (UTC)

The data used by both are the same, aren't they? I think any critique of Lynn's data collections follows both books, don't you agree? --JereKrischel 08:26, 21 January 2007 (UTC)

(1) the data are overlapping, but non-identical. There are many times more studies in RDI. (2) No, but reliability criticisms would naturally follow. Representativeness criticisms are those that say the samples from IQatWoN is not representative of the country from which it is a sample, and thus those criticisms do not naturally translate when the samples are used to estimate the IQ of the race to which the sampled individuals belong. For example, this would follow if for no other reason than that Lynn enumerates far fewer races than countries. If the criticism is that the data is just wrong to begin with, or that Lynn is a racist, then that does obviously translate. --W. D. Hamilton 08:36, 21 January 2007 (UTC)

Representativeness criticisms assert that the studies Lynn cites are not representative of what he asserts them to be. So when he takes a small study in a single country, with a small sample size errors of methodology, and then claims it represents an indigenous population, and therefore a certain "race", he is being criticized. --JereKrischel 09:16, 21 January 2007 (UTC)

You simply cannot translate a claim that the studies are unrepresentative of a nation into a claim that they are unrepresentative of a race/continent. It's too much of a leap to make without a publication to back it up. --W. D. Hamilton 09:20, 21 January 2007 (UTC)

Are you claiming that flawed data can be masked simply by increasing the number of data sets you use? I think it's quite clear that by Lynn's own admission, he has expanded upon his IQatWoN data, not filtered, replaced or revisited the claims of its unrepresentative nature. The criticism seems just as valid if he's using much of the same data. --JereKrischel 10:32, 22 January 2007 (UTC)

Historical Context

It is quite interesting that in a society where racism still persists in both subtile and blatant ways in many places one must use only "accepted channels" of research to support views that either highlight the existence of racism, or views call in to question racist notions: such as the idea that a huge, poorly defined, homogenous group of people are somehow inferior to another huge, homogenous, poorly defined group of people. And despite all nobel aspiration of 'research' that is what this topic is really about. Racism has existed clearly in "genetic" research in the past. The claim that modern research could not possibly be subject to such bias is naive. That all said, I do think that some people are working honestly to sort this question out. However, without proper historical context all of these articles carry far more certainty and credibility than they merit. This article ought to explore some of the reasons the concept of race was invented in the first place, the people who invented and the purpose it served. It should show a complete history of the idea of linking race and intelligence from the very beginning. futurebird 08:42, 21 January 2007 (UTC)

that would make an interesting subarticle which would have to grow out of the existing history section. however, most science articles do not devote the kind of space you are suggesting to an indepth history, and so neither should this one. exploring various views on race in more than a summary fashion is outside the purview of this article, but is taken up by the race article. --W. D. Hamilton 08:54, 21 January 2007 (UTC)

This isn't a scientific journal. It's an encyclopedia. --futurebird 09:34, 21 January 2007 (UTC)

But the topic is about a science. Compare with Evolution. --W. D. Hamilton 10:24, 21 January 2007 (UTC)

Evolution was never used to justify the murder and oppression of vast groups of people. futurebird 10:29, 21 January 2007 (UTC)

Nazis, social Darwinism, eugenics, etc. Notably, they overlap with this topic. --W. D. Hamilton 10:34, 21 January 2007 (UTC)

I think one of the major problems with this article is the way it has turned into a virtual journal article, exposing detail, argument and counter argument to the nth degree, in such a way that just doesn't seem encyclopedic. It may be more useful to prune out the punch/counter-punch sections, and add in a historical context section that gives a good overview (see Lieberman's table on the changing hierarchies of races for an example of a good historical brief). --JereKrischel 09:05, 21 January 2007 (UTC)

What you're describing is in part a failure to fully apply summary style, but changing the topic of the article from science to history isn't a solution and isn't tenable given the lay of the publication landscape on this topic. We write to reflect the literature. --W. D. Hamilton 09:23, 21 January 2007 (UTC)

It seems you're writing to reflect the literature that supports your particular POV. Arguably, you've created a wonderful paper worthy of inclusion into a science journal, but you've made a very poor encyclopedic article. I'd be more than willing to help you setup a mediawiki installation for your gnxp domain so that you could pursue your OR in a more appropriate venue. --JereKrischel 10:29, 22 January 2007 (UTC)

It looks like there has already been a start on gnxpwiki.com, I suggest, WDHamilton, that this article, and your extensive original research, be moved to your own wiki, and put as an external link. This article should be refactored to be more encyclopedic. --JereKrischel 09:30, 23 January 2007 (UTC)

Is there a race and musical ability article?

I've often heard it suggested that "black" people possess a better sense of rhythm (or melody?). Anyone know if any actual research has been put into this theory? --Rebroad 23:08, 21 January 2007 (UTC)

New study on relation between skin tone and income

On average, being one shade lighter has about the same effect as having an additional year of education," Hersch said.

...Hersch took into consideration other factors that could affect wages, such as English-language proficiency, education, occupation, race or country of origin, and found that skin tone still seemed to make a difference in earnings.

That means that if two similar immigrants from Bangladesh, for example, came to the United States at the same time, with the same occupation and ability to speak English, the lighter-skinned immigrant would make more money on average.... Although many cultures show a bias toward lighter skin, Hersch said her analysis shows that the skin-color advantage was not due to preferential treatment for light-skinned people in their country of origin. The bias, she said, occurs in the U.S.

http://news.yahoo.com/s/ap/20070127/ap_on_re_us/skin_tone_income

Another source for "environmental" impacts. I would like to see this brought up in relation to the other information on income. Of course, the article is still locked... futurebird 21:17, 27 January 2007 (UTC)

Grammatical Error

The line:
"Tucker have argued that some...."
should read:
"Tucker has argued that some...."

There should also be a period before the start of this sentence.
--137.99.174.125 18:45, 27 January 2007 (UTC)