The wealth of nations is mapped by their IQ

[hitssquad]

Volken's phunny phactor rotations

Originally posted by Nereid
Perhaps we don't have the same Volken paper? In the first part, Volken looks at the relative strength of three factors, assumed to be independent (i.e. an unbiased starting point), and finds: "While the total amount of variance explained by all three variables amounts to 63 percent, only 23 percent is due to the independent influence of national IQ. The remaining 40 percent, or roughly two thirds of the total variance, comes into existence due to the independent effects of economic freedom (29 percent of explained variance) and the level of democratization (11 percent of explained variance)."

The key phrase here is "the independent effects of [EF and ID]." This calculation doesn't work the way Volken says it works unless he qualifies his methods by assuming that EF and ID are not broadly influenced by national IQ (e.g., calls EF and ID "independent effects"). A major theme in Lynn's and Vanhanen's work is that a massively broad influence of national IQ -- or anything else -- will not be visible unless factors that might otherwise be assumed to be independent are put to the test as possibly non-independent factors. So, again, the Lynn/Vanhanen thesis is contingent upon the g-nexus effect generalizing from within the United States to outside the United States and therefore contributing broadly to variance in social factors outside the United States. And the fact that Volken takes issue with the generalizability of the g-nexus is not contested. Volken clearly takes issue with it -- as instanced by his assumption above of independence of factors -- and that one thing is the major point of his paper.

Volken's conclusion of "only 23 percent is due to the independent influence of national IQ" is a qualified conclusion. It would be like assuming that the 130-point average IQ of lawyers is largely independent of their positions as lawyers and therefore explains very little -- perhaps in the way of their lifestyle prudence or their investment savvy -- in terms of their incomes, when in fact factor analysis shows that lawyers do not become lawyers in the first place unless their IQs are high enough to allow them to pass the barrister's exams. Again, Volken is simply saying he takes issue with the generalizability of the g-nexus as a force of broad influence upon other factors such as EF and ID. He even states, point blank, his bias at the end of the section quoted above with the loaded phrase neglects the influence structure: "The conclusion of Lynn/Vanhanen ... is therefore fundamentally wrong, since it completely neglects the influence structure of the variables involved."

What is wrong with what Volken is saying here is that just because it is a demonstrable fact that factors can be rotated -- and, yes, they always can be rotated, depending upon the whim of the researcher or his assumption of direction of influence ("influence structure" in Volken's lexiphanicism) -- in such ways so as to make general factors appear very small, neither ipso-facto makes general factors irrelevant in specific cases, nor in general. If that were the case, then at some times a given factor would be relevant and at other times it wouldn't, depending upon who performed the latest factor rotation and why. But that is not the case. General factors are relevant to the degree that they can be maximized by factor rotation, and there is only one number (give or take a few thousandths of a degree of coefficient of congruence) in any given data set that can result when this is done. IOW, general factors are always there. You oftentimes must rotate the factors, however, in order to mathematically see them.

Jensen explains factor rotation in The g Factor.
http://www.questia.com/PM.qst?a=o&d=24373874

---
HOW INVARIANT IS g ACROSS DIFFERENT METHODS OF FACTOR ANALYSIS?

This is one of the crucially important questions in our present inquiry. Obviously the simplest way to answer it is to simulate a variety of correlation matrices that are similar to those found for actual mental test data but for which we already know the true factor structure exactly, and then see how accurately different factor analytic models and methods can estimate 7 the "true" factors known to exist in these artificial matrices.

This is just what I did, in collaboration with Dr. Li-Jen Weng, at that time a postdoctoral research scholar at the University of California, Berkeley, and a specialist in factor analysis and mathematical statistics...

Of course, we used no type of factor analysis that is expressly designed to preclude the appearance of a general factor (such as orthogonal rotation of the primary factors). We were concerned here exclusively with the amount of variation in the g factor when it is extracted by the various methods most commonly described in modern textbooks of factor analysis...

The result of this analysis was that every one of the methods of factor analysis estimated the true g so closely that there was hardly any basis for choosing between them. The congruence coefficients between the true g factor and the g factor obtained by the various methods ranged from +997 to +.999, with an average of +.998. This is especially remarkable because some of the artificial matrices were specifically designed to "trick" particular methods into yielding estimates that would deviate markedly from the true values, for example by simulating tests of highly mixed factor composition (e.g., each test having substantial loadings on all of the primary factors)...
---
(The g Factor. p81-82)



---
This orthogonal simple structure model, it turns out, has proved inappropriate in the abilities domain, and in fact Thurstone (1931) himself early on used oblique rotation of the factor axes to achieve the best possible approximation to simple structure. (Oblique factors are correlated with each other.) He only subsequently advanced orthogonal rotation to avoid some of the complications associated with oblique rotation. But it was apparent that, in the abilities domain, a good fit of the data to a simple structure model could not be achieved with orthogonal rotation, because a general factor permeates all of the primary abilities. Orthogonal rotation would achieve simple structure only if Thurstone's original theory were true. (That is the theory that mental ability consists of a number of distinct, uncorrelated abilities represented by the primary factors, and that there is no general factor in the abilities domain.) But that theory has long since been proven false. Thurstone assiduously attempted to devise tests that would provide factor-pure measures of what he called the primary mental abilities revealed by his method of multiple factor analysis. 3 But it proved impossible to devise a test that was a pure measure of any primary factor.
---
(The g Factor. p76-77)


We might note that this second instance is analogous to the Volken instance. Substitute mental ability in the above with sociological outcome, and abilities with sociological factors and we have, "That is the theory that sociological outcome consists of a number of distinct, uncorrelated sociological factors represented by the primary factors, and that there is no general factor in the sociological factors domain."



---
Any form of factor analysis that allows the extraction of a general factor has no trouble finding a very robust g in any sizable collection of Guilford's tests despite their assignment to distinct cells of the SOI. Guilford nonetheless argued that the 150 cells were orthogonal, or uncorrelated, primary factors. His empirical demonstration of so many orthogonal factors, however, relied on a technique known as targeted orthogonal rotation. Aptly named Procrustes, this method literally forces tests that were specifically selected or designed to measure the
SOI abilities to have significant loadings only on particular factors, the number and definitions of which are predetermined by the SOI model. This cannot be accepted as evidence that the 150 abilities in different cells of the SOI are not intercorrelated, since Guilford's Procrustes method of orthogonal rotation foreordains uncorrelated factors. In brief, Guilford simply assumed a priori that g does not exist, and he eschewed any type of factor analysis that would allow g to appear.
---
(The g Factor. p117)


This third instance is, again, analogous to what Volken did in his respective factor analysis of sociological factors, especially the last sentence wherein it might be instructive to simply replace Guilford's name with Volken's.

In brief, Volken simply assumed a priori that g does not exist, and he eschewed any type of factor analysis that would allow g to appear.





-Chris

 

[selfAdjoint]

Jensen's absolutely correct strictures against captious rotations have nothing to do with the selection of an explanatory model. We are not then explaining the same data by some labored trick, but rather introducing new data. Or at least unconsidered data.

There is an established procedure for establishing how much influence each of several correlates contribute, and IMHO Volken has applied it. If he had more data he could perhaps do what Herrnstein did in the Bell Curve and control for the variables separately. But his procedure here is not just obfuscation, as Gardner's was, but good statistics.

 

[hitssquad]

Volken's case for economic power of credentialism

Originally posted by selfAdjoint
We are not then explaining the same data by some labored trick, but rather introducing new data. Or at least unconsidered data.

Whether or not the Volken-introduced data should be considered as appropriate in this particular causal model is the whole point. Volken says over and over again that:

educational attainment and the education completed are not determined by IQ alone. Instead they are heavily dependent on motivation and the opportunity structure.
http://www.suz.unizh.ch/volken/pdfs/IQWealthNation.pdf
(p11)


His thesis is that opportunity structure overall, and educational opportunity in particular, is more of a factor and less of an effect. In order to be able to legitimately add educational opportunity to the Lynn/Vanhanen analysis, he has to make this case that educational opportunity is more of a factor, because by testing for any highly-loaded effect of a major factor (such as national IQ is aaumed by Lynn and Vanhanen to be), you can make that major factor appear to disappear as a causal agent -- to the degree that the effect is loaded on that major factor. The only thing Volken really accomplished was to demonstrate that IQ and educational opportunity are highly correlated. Since this very assertion of IQ/opportunity-structure linkage was itself part of Lynn and Vanhanen's thesis, Volken has brought nothing new to the table except multiple arguments about why national-IQ might not educational opportunity.

Throughout his essay Volken consistently and repeatedly commits the equivocation fallacy by equivocating national-IQ effects on education attainment with individual effects on education attainment. These two things are not necessarily equivocable because a nation providing or not providing education structure for an individual with a given IQ might make a large difference whether that person actually attains education. But to back up his decision to assume educational opportunity is an independent factor and therefore OK to include as an effector in an analysis of effects on national SES that includes IQ, Volken cites studies of documenting the influence of IQ on individuals' education attainment:

Sauer and Gattringer (1985) find that variations in educational attainment are equally determined by the child's cognitive capacity and the parent's educational aspiration for the child.
(p11)


In addition, and appropos to the last part of the above excerpt, parents' educational aspirations for the child may vary by variance in both or either the parents' IQ and the community's IQ. This is a negative contingency upon which Volken's analysis rests. Volken's calculations do not by themselves carry his case of low influence of national IQ, and Volken has to show that this assumption of his is consistent in order to justify his inclusion of educational opportunity as a factor, rather than as an effect.



And Bornschier (1988) is able to 11 demonstrate fundamental shifts in the educational opportunity structure, which promoted mass education and increasing levels of schooling.
(p11)

Variance in this educational opportunity structure may itself be largely effected by variance in national IQ. In fact, the Lynn and Vanhanen thesis rests partly upon the contingency that this is so. Therefore, Volken's thesis rests partly upon the opposite contingency that variance in IQ does not have a larger effect on educational opportunity than educational opportunity has on national income.


Volken has to show this in order to substantiate his decision to use in his analysis educational opportunity as an independent variable and not an effect. His calculations are not sufficient to justify inclusion in them of what may be a highly correlated effect of one of the major causal factors. To justify this decision, Volken uses arguments such as:


The latter is especially important, since high levels of IQ are by no means sufficient to unleash economically productive potentials. People must be adequately trained in order to fulfill complex tasks. Thus the presence of an educational system, which acts as an opportunity structure for individuals, is a necessary condition for economic development and growth. Only then can the potential of high IQ be transformed into human capital.

These are simply unsubtantiated assertions, and they fly in the face of the opposite conclusions of minimal importance to economic production of education credentials compared with IQ reached in Arthur Jensen's The g Factor.

---
Residual Validity of Amount of Education. Some employers use number of years of education or other educational credentials as a basis for selecting workers. These measures are usually valid predictors, though seldom as valid as tests of general ability, except for a specialized job where specific educational qualifications are intrinsic and essential. Educational credentials derive almost all of their predictive validity from their substantial correlation with g.
---
(p291)
http://www.questia.com/PM.qst?a=o&d=24373874


---
Lloyd Humphreys [10a] coined the term inadequate learning syndrome (ILS) to describe deficits in basic intellectual skills and information. He believes ILS is a social epidemic "as serious in its way as the AIDS epidemic." ILS is primarily a result of an insufficient level of g and is seen in the presence of adequate educational opportunity. This is what makes ILS so visible. The adverse consequence of ILS in the nation's work force is not a result of any marked change in the total population distribution of g. It is a product of the increasing demand for formal educational credentials in today's job market. As such credentials and diplomas have become spread over a greatly increased range of individual differences in actual educational achievements and qualifications, many employers have found today's high school diploma, or even a college degree, of little value
---
(p555)
http://www.questia.com/PM.qst?a=o&d=24373874



So, as I said before, what we are ultimately left with is a contingency on the Lynn/Vanhanen side that the Jensen/g-nexus effects can be generalized beyond the borders of the United States, and on Volken's side that they cannot. Lynn/Vanhanen's is the g-nexus side, and Volken's is the credentialism side. Volken has to make his case for credentialism weilding substantial economic power outside the United States, though it has already been shown not to wield substantial economic power within the United States, or his paper, for being inconsistent, is disqualified as potentially sustainable.




-Chris

 

[Nereid]

Nachtwolf: "Lynn didn't correct for the Flynn Effect, just for some 'secular trend!"
"Um, 'the secular trend' is another name for 'The Flynn Effect.'"

So the Flynn effect is the only secular trend? All observed secular trends are 'the Flynn effect'? Any new secular trend discovered will also be the Flynn effect?

Nachtwolf: I just wrote 70% to be conservative.

So, just like hitssquad, you occasionally mis-state something; I'll try to keep that in mind.

Nachtwolf: I can't wait to see what Nereid is going to post next!

No gold stars for Mark's performance in class today - she'll probably keep banging on about all the questions she posted on their own assertions which hitssquad (and Nachtwolf) have yet to answer, and follow through on flaws, inconsistencies, and irrelevancies in hitssquad's (and Nachtwolf's) proposals - i.e. more places where Nachtwolf is wrong (again), irrelevant (again), or makes unsubstantiated claims dressed up as 'facts' (again).

 

[Nereid]

Lynn's work falsifies Jensen's g nexus

hitssquad wrote: The Jensen/g nexus effects that have been established as consistently operating within the United States depend upon the presence of certain environmental and cultural conditions in order to work. If they are to work outside of the United States, then there needs to be an attending lack of a veritable environmental-cultural vacuum outside of the United States. The Lynn-Vanhanen case of generalizability of the Jensen/g-factor effects rests

1. upon a lack of consistent evidence that environmental and cultural conditions sufficient to sustain the Jensen/g-nexus effects do not generally pertain throughout the populated regions of the world;

2. upon the presence of consistent evidence that sufficient co-factors do exist such as to consistently explain anomalies found when the Jensen/g-nexus theory is generalized outside of the United States;

and

3. upon the presence of consistent evidence that these co-factors do not overwhelm the explanatory power of the Jensen/g-nexus theory when it is generalized outside of the United States.

New evidence and/or more-sophisticated analyses, could, of course, disprove the IQ and the Wealth of Nations theory. But then, that falsifiablity is partially what makes it a legitimate (from within a statistical worldview) theory.

hitssquad's summary speaks to accounting for Lynn and Vanhansen's results with explanations other than Jensen's g-nexus theory.

A simpler approach to determining its generalisability would be to first see if the L+V results themselves are consistent with Jensen's idea.

It would appear that Lynn and Vanhansen's work falsifies the Jensen idea. How?

First, an over-simplified summary of Jensen: in the US, blacks and whites have different mean IQs; the mean IQ difference is due to the different genetic makeup of blacks and whites.

Next, Jensen's own generalisation, also likely oversimplified: the mean IQ of any group is simply the weighted mean of the mean IQs of the races which comprise the group.

Third, the races of the world are the population groups identified by Cavalli-Sforza (Jensen again, also over-simplified).

Fourth, genetic distance between the races is well illustrated by the Cavalli-Sforza 2D 1st vs 2nd PC diagram (is this also Jensen? or just hitssquad?).

Conclusions consistent with the four points above:
1) the mean IQs of groups of sub-Saharan Africans will be very similar, if not the same. Why? They're all blacks, and the constituent population groups are tightly clustered.
2) the mean IQs of groups of 'Chinese' will be very similar, if not the same. Why? All 'Chinese' belong to a single population group
3) the mean IQs of groups of Europeans will very similar, if not the same. Why? They're overwhelmingly whites, and the constituent population groups are tightly clustered.

Accepting for the moment Lynn and Vanhansen's "National IQs", we see that:
1) the mean IQs are not similar; they range from 62 to 78
2) the mean IQs are not similar; they range from 98 to 110
3) the mean IQs are not similar; they range from 87 (84 including Iran) to 105.
Further, Lynn and Vanhansen claim there's little 'noise' in their data (although they don't seem to have given figures for relative or absolute accuracy); +/- 2 would seem reasonable.

Conclusion: data inconsistent with hypothesis; hypothesis falsified.

 

[hitssquad]

The Flynn effect as a secular trend

Originally posted by Nereid
Flynn effect is the only secular trend?

No. The Flynn effect is the secular trend of rising population IQ-test scores over time, or "the increase in raw scores on various IQ tests in many populations over the last sixty years or so," as Jensen says below.

---
THE SECULAR INCREASE IN IQ

One of the most puzzling phenomena is the increase in raw scores on various IQ tests in many populations over the last sixty years or so. This phenomenon has been under investigation since the mid-1980s. Most of the evidence for the upward trend comes from the many past studies where various tests that were originally normed at one time on a representative population sample were renormed many years later on a different, but supposedly equivalent, population sample.

This upward trend in the population's mean test scores has been aptly dubbed the "Flynn effect," after James R. Flynn, a professor of political science at the Otago University in New Zealand, who was responsible for amassing most of the evidence for what he has referred to as "massive IQ gains."[ 11 ] The bulk of this evidence comes from the period between 1930 and 1980.

Before summarizing this evidence, it should be noted that when a test is normed or renormed, the IQ (which is a standardized score) is always scaled such that the population mean is 100 and the standard deviation is fifteen. Population trends in actual test performance, as indicated by raw scores (number right), therefore, are not reflected by the IQ, except to some degree as the trend proceeds between the original norming and the renorming of the same test. Actual gains in test performance over long periods are adequately measured only by raw scores.

In measuring these raw-score gains two problems must be considered: (1) The change in raw scores must be demonstrated on the identical test administered on both occasions. Changes in test items (e.g., dropping some old items and substituting new ones) may alter the overall difficulty level of the test, causing a spurious rise (or fall) in the mean raw score of the more recently tested sample; (2) a much more problematic condition in renorming tests (or in comparing the same test on different samples that were tested at widely separated times) is the assumption that the two norm samples are truly equivalent. A number of factors militate against obtaining equivalent and representative samples of a population. The most obvious are population changes over decades or generations, due to changing demographics, such as birth rates in different socioeconomic segments of the population, rates of immigration and emigration, regional changes in the types of employment available, and the like.

Although the supposed equivalence of samples taken at different times is often open to doubt, changing demographics should not cause changes in test scores that are consistently in one direction for every test in every study conducted with many different population samples. Flynn's compilations of changes in test scores over decades and generations were drawn from fifteen economically advanced nations in North America, Europe, and Asia. In addition, since the publication of Flynn's major reviews, other investigators have reported highly similar results based on data from other countries and on tests not included in Flynn's reviews. The overwhelming consistency of virtually all of the data with respect to the direction of the trend in test scores leaves little doubt of the reality of the "Flynn effect." Whatever inconsistencies exist are all in the details.



11. Flynn, 1984, 1987a, 1994.


Flynn J. R. ( 1980). Race, IQ and Jensen. London: Routledge & Kegan Paul.

Flynn J. R. ( 1984). "The mean IQ of Americans: Massive gains 1932 to 1978". Psychological Bulletin, 95, 29-51.

Flynn J. R. ( 1987a). "Massive gains in 14 nations: What IQ tests really measure". Psychological Bulletin, 101, 171-191.

Flynn J. R. ( 1987b). Race and IQ: Jensen's case refuted. In S. Modgil & C. Modgil (Eds.), Arthur Jensen: Consensus and controversy (pp. 221-232). New York: Falmer.

Flynn J. R. ( 1987c). "The ontology of intelligence". In J. Forge (Ed.), Measurement, realism and objectivity (pp. 1-40). New York: D. Reidel.

Flynn J. R. ( 1990). "Massive IQ gains on the Scottish WISC: Evidence against Brand et al.'s hypothesis". Irish Journal of Psychology, 11, 41-51.

Flynn J. R. ( 1991). Asian Americans: Achievement beyond IQ. Hillsdale, NJ: Erlbaum.

Flynn J. R. ( 1993). "Skodak and Skeels: The inflated mother-child gap". Intelligence, 17, 557-561.

Flynn J. R. ( 1994). "IQ gains over time". In R. J. Sternberg (Ed.), Encyclopedia of human intelligence (pp. 617-623). New York: Macmillan.

Flynn J. R. ( 1996). "What environmental factors affect intelligence: The relevance of IQ gains over time"
---
(AR Jensen. The g Factor. p318-319.)
http://www.questia.com/PM.qst?a=o&d=24373874





-Chris

 

[Nereid]

forget IQ; nothing beats city living

This website has nice pop-up tables, in bite-sized chunks, suitable for toy researchers and others:
http://www.carleton.ca/cifp/rank.htm

One of the tables you can pull up is the 'urbanisation rank score' of ~180 countries, which is a handy little demographic indicator of the extent to which a country's population is concentrated in urban centres.

If you do a simple regression analysis of these rank scores against the real per capita GDP which L+V used, you find there's a nice correlation, of ~73%. Now as Nachtwolf has so kindly informed us, "Correlations over 50% don't just drop out of the sky". If you then wonder how much variation is left over for L+V's "National IQ", well some, but not much; it has a ~38% correlation. (I didn't run a multiple regression analysis). And this is only a rank score! (mostly integers, ranging from 1 to 9) Imagine if there were some more nuanced metric.

And the outliers? Well, the US and Iraq - the former is too wealthy for its urbanisation rank score of 8, and Iraq is far too poor. Interestingly, South Korea isn't really exceptional this time, and Singapore's and Hong Kong's smarts are all down to them being all city.

So, what can we conclude? That as a country becomes more urbanised its "National IQ" rises? Well, gee whizz, that sounds like it might explain the Flynn effect too

Seriously folks, how many other high-correlation, non-IQ factors are lurking out there? Maybe geography and economics are all you need.

 

[hitssquad]

Originally posted by Nereid
a) this particular classification of races is L. L. Cavalli-Sforza, P. Menozzi P. & A. Piazza's, per their 1994 book "The history and geography of human genes"

c) the "east Asian" races are (reading down the tree) Mongol, Tibetan, Korean, Japanese, Ainu, [gap], S. Chinese, Mon Khmer, Thai, Indonesian, Philippine, Malaysian

No. Generally, there is a distinction drawn between Northeast Asians (mongoloids; Han) such as Han Chinese, Korean and Japanese and South Asians/Pacific Islanders such as Thai and Philippine. As can be seen on the genetic linkage tree, caucasians are more-closely related to Northeast Asians than Northeast Asians are related to South Asians/Pacific Islanders.

--

                           |---------------------San (Bushmen)
 __________________________|    |----------------Mbuti Pygmy
|                          |____|          __|---Bantu
|                               |         |  |---Nilotic
|                               |         |------W. African
|                               |----------------Ethiopian
|                                                
|                                   |------------S.E. Indian
|                       |-----------| |----------Lapp
|                       |           |-||---------Berber, N. African
|                       |              | |-------Sardinian
|                       |              |_||------Indian
|                       |                ||      
|                       |                || _|---S.W. Asian
|                       |                ¯|| |_|-Iranian
|                   ____|                 -|   |-Greek
|                  |    |                  |-|---Basque
|                  |    |                    ¯|--Italian
|                  |    |                     ¯|-Danish
|                  |    |                      |-English
|                  |    |                        
|                  |    |             _____|-----Samoyed
|             |----|    |            |     |-----Mongol
|             |    |    |       |----|  __|------Tibetan
|             |    |    |       |    | |  |__|---Korean
|             |    |    |       |    |-|     |---Japanese
|             |    |    |  |----|      |---------Ainu
|             |    |    |  |    |                
|-------------|    |    |  |    |--|-------------N. Turkic
              |    |    |  |       |____|--------Eskimo
              |    |    |  |            |--------Chukchi
              |    |    |--|                __|--S. Amerind
              |    |       |          |----|  |--C. Amerind
              |    |       |----------|    |-----N. Amerind
              |    |                  |----------N.W. American
              |    |                             
              |    |                  _____|-----S. Chinese
              |    |                _|     |__|--Mon Khmer
              |    |               | |        |--Thai
              |    |          _____| |_|---------Indonesian
              |    |         |     |   |---------Philippine
              |    |---------|     |-------------Malaysian
              |              |___|---------------Polynesian
              |                  |__|------------Micronesian
              |                     |------------Melanesian
              |________|-------------------------New Guinian
              |        |-------------------------Australian

--
http://f1.pg.photos.yahoo.com/ph/hitssquad/album?.dir=/ee6c

d) apart from small representations of groups such as Russian and Polynesian, the >50 ethnic groups in China ("nationalities" as they refer to themselves) are various mixtures of Mongol, Tibetan, Korean, S. Chinese, Mon Khmer (maybe), and Thai (maybe).

China's population is mostly Han, especially in the Northeast. This site says China is 92% Han:
http://www.travelchinaguide.com/intro/nationality/

And reiterates that non-Han Chinese do not tend to live in the Northeast:

--
Minority Ethnic Groups

Most of these 7 percent live in the vast areas of the West, Southwest and Northwest.
--

How do Cavalli-Sforza et al's population groups correlate with the results from the many studies into variation within the human genome?

Jensen says Cavalli-Sforza et al (1994) is "so highly correlated" with Nei & Roychoudhury (1993) "as to be virtually equivalent for most purposes."

Could you please briefly summarise the research technique that Jensen used to determine a) the extent to which each of Cavalli-Sforza's population groups ('races') was a 'breeding population'

AFAIK, Jensen did not confirm that they were actually having sex. He did confirm Cavalli-Sforza's results of analysis of distributions of genetic markers, and his results were...

--
(1) Mongoloids, (2) Caucasoids, (3) South Asians and Pacific Islanders, (4) Negroids, (5) North and South Amerindians and Eskimos, (6) aboriginal Australians and Papuan New Guineans.
--
(p518)

...which are essentially the same as Cavalli-Sforza's results:
http://f1.pg.photos.yahoo.com/ph/hitssquad/detail?.dir=/ee6c&.dnm=25d6.jpg

fuzziness of the boundaries?

IIRC Jensen did not quantify the fuzziness of the boundaries. He did, however, quantify population mean variances. He did this by computing the variance ratio of the population's phenotypic characteristics, which is the variance in phenotypic characteristics between populations divided by the variance in characteristics within populations:

--
4. One often hears it said that the genetic differences within racial groups (defined as statistically different breeding populations) is much greater than the differences between racial groups. This is true, however, only if one is comparing the range of individual differences on a given characteristic (or on a number of characteristics) within each population with the range of the differences that exist between the means of each of the separate populations on the given characteristic. In fact, if the differences between the means of the various populations were not larger than the mean difference between individuals within each population, it would be impossible to distinguish different populations statistically. Thinking statistically in terms of the analysis of variance, if we obtained a very large random sample of the world's population and computed the total variance (i.e., the total sum of squares based on individuals) of a given genetic character, we would find that about 85 percent of the total genetic variance exists within the several major racial populations and 15 percent exists between these populations. But when we then divide the sum of squares (SS) between populations by its degrees of freedom to obtain the mean square (MS) and we do the same for the sum of squares within populations, the ratio of the two mean squares, i.e., Between MS/Within MS, (known as the variance ratio, or F ratio, named for its inventor, R. A. Fisher) would be an extremely large value and, of course, would be highly significant statistically, thus confirming the population differences as an objective reality.
--
(p516-517)
http://www.questia.com/PM.qst?a=o&d=24373874



Here are some of the phenotypic characteristics he used to quantify population mean variances:

--
5. Among the genetically conditioned physical differences in central tendency, nearly all attributable to natural selection, that exist between various contemporary breeding populations in the world are: pigmentation of skin, hair, and eyes, body size and proportions, endocranial capacity, brain size, cephalic index (100 × head-width/head-length), number of vertebrae and many other skeletal features, bone density, hair form and distribution, size and shape of genitalia and breasts, testosterone level, various facial features, interpupillary distance, visual and auditory acuity, color blindness, myopia (nearsightedness), number and shape of teeth, fissural patterns on the surfaces of teeth, age at eruption of permanent teeth, consistency of ear wax, blood groups, blood pressure, basal metabolic rate, finger and palm prints, number and distribution of sweat glands, galvanic skin resistance, body odor, body temperature, heat and cold tolerance, length of gestation period, male/female birth ratio, frequency of dizygotic twin births, degree of physical maturity at birth, physical maturation rate, rate of development of alpha (brain) waves in infancy, congenital anomalies, milk intolerance (after childhood), chronic and genetic diseases, resistance to infectious diseases ( Baker, 1974; Harrison et al., 1964; Rushton, 1995). Modern medicine has recognized the importance of racial differences in many physical characteristics and in susceptibilities to various diseases, chronic disorders, birth defects, and the effective dosage for specific drugs. There are textbooks that deal entirely with the implications of racial differences for medical practice ( Lin et al., 1993; Nesse & Williams, 1994; Polednak, 1989). Forensic pathologists also make extensive use of racial characteristics for identifying skeletal remains, body parts, hair, blood stains, etc.
--
(p517)
http://www.questia.com/PM.qst?a=o&d=24373874

[Edit: as SelfAdjoint points out, Cavalli-Sforza didn't use the concept of 'race', so asking about how the groups he identified match hitssquad's definition of race is silly

Using a word and using a concept are different things. Cavalli-Sforza used the concept of race.




-Chris

 

[Nereid]

hitssquad: Generally, there is a distinction drawn between Northeast Asians (mongoloids; Han) such as Han Chinese, Korean and Japanese and South Asians/Pacific Islanders such as Thai and Philippine. As can be seen on the genetic linkage tree, caucasians are more-closely related to Northeast Asians than Northeast Asians are related to South Asians/Pacific Islanders.

I'm having trouble finding 'Han Chinese' in the diagram; the NE Asians are Samoyed (who live primarily in Russian's far east?), Mongol, Tibetan (there are a lot of both groups in China's west and north-west, particularly the Tibetan Autonomous Region, the Inner Mongolian Autonomous Region, and Qinghai), Korean (some of whom live in China's north east), Japanese, and Ainu (who are a repressed minority in Japan?). The few hundred 'Russians' in China tend to live in the northeast, esp in Harbin.

As the travel guide says, the Zhuang are the biggest non-Han group, and they have their own Autonomous Region too - Guangxi, which is next door to Guangdong, whose capital used to be called Canton (now Guangzhou).

hitssquad: Jensen says Cavalli-Sforza et al (1994) is "so highly correlated" with Nei & Roychoudhury (1993) "as to be virtually equivalent for most purposes."

Curious that the biggest single population group by far on this planet isn't in C-S's diagram (at least, I can't see it). But then, Burma/Myanmar also has >50 ethnic groups, and since C-S's diagram has only 42 groups in total, I guess there's plenty of work for young PhDs.

AFAIK, detailed studies of variation in the human genome are being undertaken with a view to being able to more confidently prescribe medicines and get a 'heads up' on possible side-effects (among other things), e.g. the studies being funded/coordinated by the Wellcome Trust. The early results have been quite interesting, and produced not a few surprises (it would seem there's been an awful lot more inter-breeding going on than Jensen et al had supposed), and exploded many a myth (e.g. lactose intolerance of Han Chinese; guess which large market for yoghurt products is growing the fastest?). Maybe the g-nexus ideas too will be shown to be hopelessly simplistic before long?

 

[hitssquad]

Originally posted by Nereid
how many other high-correlation, non-IQ factors are lurking out there?

The entire g nexus.

And I already answered this, in another way, here:
https://www.physicsforums.com/showthread.php?s=&postid=144471#post144471


In addition, Lynn answered it, here:

--
It has been pointed out that correlation analysis does not establish causality because of the fact that correlations merely measure covariation. Let us consider what causality presupposes. Manheim and Rich (1986: 21-22) say that it is justified to postulate causal relationships only when four conditions are simultaneously met: First, the postulated cause and effect must change together, or covary. Second, the cause must precede the effect. Third, we must be able to identify a causal linkage between the supposed cause and effect. Fourth, the covariance of the cause and effect phenomena must not be due to their simultaneous relationship to some other third factor. We think that the relationship between national IQ and the measures of per capita income and economic growth meets these requirements quite well. First, correlations indicate that the postulated cause and effect change together. Second, because differences in national IQs are partly genetic, they have certainly preceded contemporary differences in economic conditions. Third, the causal linkage between the hypothesized cause and effect will be discussed and explained in the next section. Fourth, it is highly improbable that the observed covariance between cause and effect could be due to any third factor. This last requirement will be discussed in greater detail in the next section. Consequently, we are quite confident that the relationship is causal.

--
http://www.rlynn.co.uk/pages/article_intelligence/5.htm




--
There are two reasons why we consider that a causal effect of national IQ on per capita incomes and rates of economic growth is the most reasonable theory to explain the correlations. First, this theory is a corollary of an already established body of theory and data showing that IQ is a determinant of income among individuals, the evidence for which has been reviewed in the introduction. IQs measured in childhood are strong predictors of IQs in adolescence and these are predictors of earnings in adulthood. The most reasonable interpretation of these associations is that IQ is a determinant of earnings. From this it follows that groups with high IQs would have higher average incomes than groups with low IQs because groups are aggregates of individuals. This prediction has already been confirmed in the studies of the positive relationship between IQs and per capita incomes among the American states and among the regions of the British Isles, France and Spain, as noted in the introduction. The positive relation between IQ and income is so well established that it can be designated a law, of which the finding that national IQs are positively related to national per capita incomes is a further instance.

Second, there is a straightforward explanation for the positive association between IQ and incomes at both the individual and population level. The major reason for this association is that people with high IQs can acquire complex skills that command high earnings and that cannot be acquired by those with low IQs. Nations whose populations have high IQs tend to have efficient economies at all levels from top and middle management through skilled and semi-skilled workers. These nations are able to produce competitively goods and services for which there is a strong international demand and for which there is therefore a high value, and that cannot be produced by nations whose populations have low IQs. In addition, nations whose populations have high IQs will have intelligent and efficient personnel in services and public sector employment that contributes indirectly to the strength of the economy such as teachers, doctors, scientists and a variety of public servants responsible for the running of telephones, railroads, electricity supplies and other public utilities. Finally, nations whose populations have high IQs are likely to have intelligent political leaders who manage their economies effectively. Skilled economic management is required to produce the right conditions for economic growth, such as keeping interest rates at the optimum level to produce full employment with minimum inflation, maintaining competition, preventing the growth of monopolies, controlling crime and corruption, and promoting education, literacy and numeracy and vocational training.


...it might be argued that national per capita incomes are a cause of national differences in IQs. This argument would state that rich nations provide advantageous environments to nurture the intelligence of their children in so far as they are able to provide their children with better nutrition, health care, education and whatever other environmental factors have an impact on intelligence, the nature of which is discussed in Neisser (1998). Intelligence has increased considerably in many nations during the twentieth century and there is little doubt that these increases have been brought about by environmental improvements, which have themselves occurred largely as a result of increases in per capita incomes that have enabled people to give their children better nutrition, health care, education and the like. Such a theory has some plausibility but it cannot explain the totality of the data. Countries like Japan, South Korea, Taiwan and Singapore had high IQs in the 1960s when they had quite low per capita incomes and the same is true of China today. Nevertheless, the model of national differences in IQ as a major determinant of economic growth and per capita incomes should probably be supplemented by the postulation of a small positive feedback in which national per capita income has some impact on the population's IQ.

Our results are based on a sample of 60 nations out of approximately 185 nations of significant size in the world. We believe that the sample can be regarded as relatively well representative of the totality of nations because all categories of nations are well represented including the economically developed "First World" market economies of North America, Western Europe, Australia and New Zealand; the "Second World" former communist nations of Russia and Eastern Europe; the "Third World" economically developing but impoverished nations of South Asia, sub-Saharan Africa and the Caribbean; and the residual categories of Latin America and East Asia. If the representativeness of our sample is accepted, our results indicate that slightly over half the variance in national per capita income in the contemporary world is attributable to national differences in IQ. However, it should be noted that correlations are somewhat lower in the total group of 185 countries (see Lynn and Vanhanen, 2002). The difference in correlations implies that this sample of 60 nations is probably slightly biased.

The regression analysis suggests that a major additional factor is the economic form of organisation consisting of whether countries have market or socialist economies. The countries that have the largest positive residuals and therefore have higher per capita income than would be predicted from their IQs are Australia, Belgium, Canada, Denmark, France, Ireland, Israel, Qatar, Singapore, South Africa, Switzerland and the United States. With the exception of Qatar and South Africa, all of these are technologically highly developed market economy countries and their higher than predicted per capita incomes can be attributed principally to this form of economic organisation. Qatar's exceptionally high level of per capita national income is principally due to its oil production industries. South Africa's much higher than expected level of per capita income should probably be attributed principally to the cognitive skills of its European minority who comprise 14 per cent of the population.

The countries that have the largest negative residuals are China, Iraq, South Korea, the Philippines, Romania, Russia, Slovakia, Thailand and Uruguay. Four of these countries (China, Romania, Russia and Slovakia) are present or former socialist countries whose economic development has been hampered by their socialist economic and political systems. After the collapse of the Soviet communist systems in 1991 and the introduction of market economies in these countries and in China, the prospects for rapid economic development for these countries are good, although it takes time to establish effective market economies. Of the remaining five countries with large negative residuals, Iraq's low level of per capita national income is due principally to the destruction inflicted in 1990 war and the UN sanctions imposed in 1990. South Korea's Real GDP per capita is also considerably lower than expected on the basis of the country's exceptionally high level of national IQ (106). The principal explanation for this is probably that South Korea had a very low per capita income at the end of World War Two as a result of military defeat and occupation by the Japanese and that it has not yet had sufficient time to achieve the predicted level of per capita income, although economic growth in South Korea since 1950 has been extremely high (see Appendix 2). The Asian economic crisis in 1998 may have increased the negative residuals of the Philippines and Thailand temporarily. Economic growth in Uruguay has been strong since the 1970s, although the country has not yet achieved the per capita income level expected on the basis of its relatively high national IQ.
--
http://www.rlynn.co.uk/pages/article_intelligence/6.htm





-Chris

 

[hitssquad]

The number and composition of different racial groups that can be reliably determined

Originally posted by Nereid
Burma/Myanmar also has >50 ethnic groups

--
Of course, any rule concerning the number of gene loci that must show differences in allele frequencies (or any rule concerning the average size of differences in frequency) between different breeding populations for them to be considered races is necessarily arbitrary, because the distribution of average absolute differences in allele frequencies in the world's total population is a perfectly continuous variable. Therefore, the number of different categories, or races, into which this continuum can be divided is, in principle, wholly arbitrary, depending on the degree of genetic difference a particular investigator chooses as the criterion for classification or the degree of confidence one is willing to accept with respect to correctly identifying the area of origin of one's ancestors.

Some scientists have embraced all of Homo sapiens in as few as two racial categories, while others have claimed as many as seventy. These probably represent the most extreme positions in the "lumper" and "splitter" spectrum. Logically, we could go on splitting up groups of individuals on the basis of their genetic differences until we reach each pair of monozygotic twins, which are genetically identical. But as any pair of MZ twins are always of the same sex, they of course cannot constitute a breeding population. (If hypothetically they could, the average genetic correlation between all of the offspring of any pair of MZ twins would be 2/3; the average genetic correlation between the offspring of individuals paired at random in the total population is ½; the offspring of various forms of genetic relatedness, such as cousins [a preferred match in some parts of the world], falls somewhere between 2/3 and ½.) However, as I will explain shortly, certain multivariate statistical methods can provide objective criteria for deciding on the number and composition of different racial groups that can be reliably determined by the given genetic data or that may be useful for a particular scientific purpose.
--
(p425-426)
http://www.questia.com/PM.qst?a=o&d=24373874

it would seem there's been an awful lot more inter-breeding going on than Jensen et al had supposed

In 1998, Jensen put the typical european component in American blacks at 25%.
(p432)
http://www.questia.com/PM.qst?a=o&d=24373874


Lately, there have been indications that he overestimated inter-breeding:

--
Shriver's study shows that they are less European that was previously believed.

Earlier, cruder studies, done before direct genetic testing was feasible, suggested that African-Americans were 25 or even 30 percent white. Shriver's project is not complete, but with data from 25 sites already in, he is coming up with 17-18 percent white ancestry among African-Americans. That's the equivalent of 106 of those 128 of your ancestors from seven generations ago having been Africans and 22 Europeans.
--
http://groups.yahoo.com/group/e-l/message/9213





-Chris

 

[Nereid]

Lynn - more flaws

Some interesting things about Lynn's dataset (http://www.rlynn.co.uk/pages/article_intelligence/t3.htm ):

If you divide the 'nations' into three - sub-Saharan, 1998 per capita real GDP >16k, and the rest, what do you find (in terms of the relationship between "National IQ" and wealth and its growth)?

1) there are no meaningful correlations between the two variables within the first and second groups (with one exception - see 4 below)

2) only one of the three correlations in the 'rest' group is all that noticable (1998 real per capita GDP and "National IQ", R² ~0.43)

3) South Korea is an outlier; if this single data point is removed, growth correlations in the 'rest' group all but disappear

4) Japan causes almost all the correlation in the 'rich' group.

In simple English:
1) the differences between the wealth of 'rich' countries have nothing to do with the mean IQ of their inhabitants
2) as a country gets richer, the mean IQ of its inhabitants goes up.

So what about the sub-Saharan countries? Perhaps hitssquad would be kind enough to tell us how, in detail, L+V processed the data found in the original reports they list here (http://www.rlynn.co.uk/pages/article_intelligence/7-a1.htm ) to obtain the "National IQ" numbers they use in their analyses? Such information is absent from Lynn's webpage. In particular:
a) what secular trends (Flynn effect or other) have been included?
b) how they assured themselves that the sampling done by the original authors was unbiased, that the testing protocol was uniformly equal and unbiased, and of the validity of including IQ scores of below 50 (60? 65?)* from SPM test?

What about L+V's four tests (does correlation mean causation; hitssquad provided a quote earlier)?

"First, the postulated cause and effect must change together, or covary." L+V 1/9 (covariance is confined to one subset of the data, and only in one of the three relationships).

"Second, the cause must precede the effect." L+V 0. Without good data on the secular trends in "National IQ" in the 19 (20) countries comprising the 'rest', precedence remains to be established.

"Third, we must be able to identify a causal linkage between the supposed cause and effect." L+V 0.
As I've said in earlier posts, L+V's proposed cause (some races have smarter genes than others) may have been established in the US, but as Jensen himself freely admits, this hasn't been established elsewhere. Further, Lynn himself admits of the possibility of the wealth causing changes in IQ: "... the model of national differences in IQ as a major determinant of economic growth and per capita incomes should probably be supplemented by the postulation of a small positive feedback in which national per capita income has some impact on the population's IQ." Finally, there's ample evidence - at a 'country' llevel - to cast serious doubt on L+V's proposed 'causal link'.

"Fourth, it is highly improbable that the observed covariance between cause and effect could be due to any third factor." L+V 0.
Since there is, in fact, very little 'observed covariance', it's very probable that any that there is could be due to a third (or fourth) factor. Any number of geographic or economic factors might be responsible.

So what's the real story on sub-Saharan countries? Perhaps Paleo would like to address this question?

*L+V: "It has been suggested by a referee that the mean IQs of sub-Saharan African countries are so low that they cannot be valid and that they spuriously inflate the correlations between the national IQs and the measures of per capita income and economic growth."
from a quote in an earlier hitssquad post (my emphasis): "The greater B-W differences on the RT and RTSD components of the ECTs in the South African study is best explained by the fact that this group of South African blacks scored, on average, about 2ó below British (or South African) whites, while there is only about 1ó difference between American blacks and whites. 59 In the Lynn and Holmshaw study, the W-B difference on Raven's Standard Progressive Matrices (SPM) was about 2.5ó. But we cannot be very confident of this value, because the SPM appeared to be too difficult for the African blacks. Their mean raw score on the SPM was only about three points above the chance guessing score, which casts doubt on the reliability and validity of the SPM as a measure of individual differences in g for this sample."

[Edit: fixed formats, some typos]

 

[Nereid]

Ireland

... has a "National IQ" of 87 (according to L+V), the lowest of the 16 (18) 'rich' countries. And yet ... its long term growth (1976-1998) is the second highest in this group of countries!

Lynn: "there is a straightforward explanation for the positive association between IQ and incomes at both the individual and population level. The major reason for this association is that people with high IQs can acquire complex skills that command high earnings and that cannot be acquired by those with low IQs. Nations whose populations have high IQs tend to have efficient economies at all levels from top and middle management through skilled and semi-skilled workers. These nations are able to produce competitively goods and services for which there is a strong international demand and for which there is therefore a high value, and that cannot be produced by nations whose populations have low IQs. In addition, nations whose populations have high IQs will have intelligent and efficient personnel in services and public sector employment that contributes indirectly to the strength of the economy such as teachers, doctors, scientists and a variety of public servants responsible for the running of telephones, railroads, electricity supplies and other public utilities. Finally, nations whose populations have high IQs are likely to have intelligent political leaders who manage their economies effectively. Skilled economic management is required to produce the right conditions for economic growth, such as keeping interest rates at the optimum level to produce full employment with minimum inflation, maintaining competition, preventing the growth of monopolies, controlling crime and corruption, and promoting education, literacy and numeracy and vocational training."

Not only has Ireland managed to do all the things which Lynn implies it can't (the Irish aren't smart enough), but it has developed an enviable leadership position in telecoms software development.

If the Irish can succeed so well with a mean IQ of 87, then 'the blacks' in the US (with a mean IQ of 85) should be able to do just as well, surely?

Ireland today, India (82), Iran (84), Cuba (85), Egypt (83), ... tomorrow?

Evidence? CMM (Capability Maturity Model) is a widely used benchmark for software development companies; level 5 is the highest. Very few companies are CMM Level 5; among them ... Wipro, an Indian company. Just in case you haven't been following: "*SNIP These nations [ones with high IQs] are able to produce competitively goods and services for which there is a strong international demand and for which there is therefore a high value, and that cannot be produced by nations whose populations have low IQs." (Lynn, my emphasis)

 

[Nereid]

'race' as used in L+V, and by Jensen

Earlier in this thread hitssquad posted links to diagrams used in Jensen's "The g Factor", which in turn were taken from Cavalli-Sforza et al. The diagrams list C-S's estimated 42 pre-1492 population groups and the genetic distances between them.

L+V also use non-national groups of people, which they usually (but not always) call 'races'. However, they do not list them on their website, and it's difficult to determine how important a role they play in their work.

Here is a list of such groups, from the Lynn website, in alphabetical order (from the context it's clear that sometimes L+V are quoting others; other times it's their own terminology):

African (black)
Arab
Blacks
Chinese
Colored
Creole (black-white hybrids)
Indians
Malagasy
Malayan-Polynesian
Malays
Mixed black-white
Negroid
Whites

What is the correspondence with C-S's list? Only those in blue are C-S population groups (note that "Malays" may not be the same as "Malaysians"). Those in green are higher-level groups.

And Jensen? After a lengthy introduction (hitssquad has posted the links), states: "Since far more empirical research relevant to the examination of the default hypothesis with respect to g has been done on the black-white difference, particularly within the United States, than on any other populations, I will focus exclusively on the causal basis of the mean black-white difference in the level of g."

Interestingly, Jensen appears to make no attempt to relate "Whites" to the ~>5 C-S population groups which may be 'white' (other than that they're 'European'). Wrt 'blacks', he mentions their predominant west African ancestry without identifying which of the C-S population groups they may have come from. He also points out that there is a significant 'white' component in their ancestry, ranging from ~4% to >40%.

 

[Nereid]

To anyone wishing to use Lynn's study

(other than to rip more holes in it, use it as a textbook study of how NOT to do science, ...); and to Nachtwolf, hitssquad, Carlos, Apollo and any others who've been keen to blow Lynn's trumpet.

Summary of questions I have asked and which have not yet been answered by Lynn's defenders:

Research and analysis:
Would someone please give us a simple summary of the steps Lynn and Vanhanen used to get from the results of a test administered to x people in country y in year t to a "National IQ" figure? I'm particularly interested in the following:
- initial sample selection
- test protocol
- analysis of test results (esp distribution about the observed mean)
- key stages in analyses of data, to get to "National IQ", especially external assumptions and inputs.

In particular:
a) what secular trends (Flynn effect or other) have been included?
b) how they assured themselves that the sampling done by the original authors was unbiased, that the testing protocol was uniformly equal and unbiased, and of the validity of including IQ scores of below 50 (60? 65?) from SPM test?

Ghana:
L+V assert that the "National IQ" of Ghana is 62.
If "IQ 70-75 [is] often [...] considered the threshold for mental retardation", I would guess that 40-50 would be the threshold for severe retardation, perhaps those with Down’s Syndrome have IQs in this range? Would people with such low IQs – of whom there must have been at least 100 in the sample which lead L+V to their determination - to be able to do CPM or SPM? Would the test protocol have to be changed to administer either test to such people? What sort of sampling technique was used to ensure that such severely retarded people (if indeed they were) could be included as test subjects?

Then the onus is on those using or defending L+V’s results support to use publicly available data to estimate the minimum IQ needed to get into the leading university in Ghana, and to show that it's consistent with a "National IQ" of 62. Further, as an independent test, such defenders should compile data on the number of Ghanans who are employed as pilots by international airlines, academics in the US/EU/Japan/Australia (etc), and other professions in countries where a priori you would expect an IQ of 100 essential to be employed. Then we can calculate the expected number of such folk, based on Ghanan demographic data, and Lynn's assertion that the "National IQ" is 62.

Statistical consistency:
1) Here are the reported results from two studies used by Lynn and Vanhanen, done using the same instrument, on samples purporting to be randomly drawn from the same population:
A: mean 103, sample size 43,825
B: mean 105, sample size 2,496

It is claimed that the difference in the population mean, inferred from these two studies, is not statistically significant. Do you agree? Explain your answer.

2) Please present for us the "National IQ" data Lynn and Vanhanen use, for China, Hong Kong, and Taiwan. Considering that the total non-Han ("Chinese" to you, I guess) population of greater China accounts for ~8% of the total population, please use L+V's data to give a consistent, statistically sound number for the IQ of the Han "race". Please draw our attention to any obvious, systematic trends in the data.

The 1984 Wall Street Journal ad, researchers in this field, etc:
1) Is this Wall Street Journal statement intended to refer to "individual and group differences in intelligence" throughout the world, or just in the US?

2) There are ~50 separate authors mentioned in the 81 studies quoted by Lynn. There are ~50 signatories to the document jerryel quoted. There are ONLY TWO names on both (and even one of those may be a coincidence) - Lynn and Gilmore. What is going on in this field?

3) Was Raven still an active academic in 1994? If so, why didn't he sign too?

4) What was the reception to Lynn's work among the hundreds of US profs who didn't sign the 1994 newspaper ad?

5) "Sir Cyril Burt", part of the "the intellectual core group of the hereditarian school of psychology" ... isn't he the guy whose work on the IQ differences among twins - especially those raised apart from birth - was later shown to be almost entirely fraudulent? To what extent has that fraudulent work (if that's what it was) continued to be used (knowingly or unknowingly; directly or indirectly) in conclusions about the hereditability of intelligence?