Wednesday, March 29, 2006

PROF. FRASER REPLIES

Various people have made complaints to Australia's "Human Rights and Equal Opportunity Commission" about statements made by Prof. Andrew Fraser of Macquarie University in Sydney. Below is the reply Prof. Fraser sent to the HREOC

My Public Comments:

The complaint lodged by Mr Hareer focuses on a letter written by me to the editor of the Parramatta Sun and published on 6 July 2005. The text of that letter was as follows:

Now that a large number of Sudanese refugees have been settled in the Parramatta-Blacktown area, Anglo-Australians are once again expected to acquiesce in the steady erosion of their distinctive national identity.

Australia, it seems, can no longer remain the homeland of a particular people. Instead, it must become a colony of the Third World.

Thirty years ago, no one in the world had any difficulty identifying an Australian. Today, if the headline in the Sun is to be believed, black Africans and Muslim Afghanis “are Aussies just like” the descendants of the Anglo-Celtic pioneers who settled and built this country.

Community Relations Commissioner Stepan Kerkyasharian declares that “Australians…have a responsibility” to help those on the losing side in Third World civil wars to settle here, wherever and whenever it suits governments and the ever-expanding refugee industry. He assures us that the ethnic and religious conflicts endemic to every other part of the world will be magically dissolved by the state-enforced “commonality of Australianism.”

That utopian fantasy is particularly likely to unravel as local African tribal groups grow in size and confidence. Experience practically everywhere in the world tells us that an expanding black population is a sure-fire recipe for increases in crime, violence and a wide range of other social problems.

The fact is that ordinary Australians are being pushed down the path to national suicide by their own political, religious and economic élites. Shutting our eyes to that fact will not make it go away.



In the course of media interviews conducted after that letter was sent to the Parramatta Sun, I expressed the view that the White Australia Policy had been based upon the perfectly sensible premise that ethnic homogeneity in any nation is a source of strength and unity. Conversely, I argued, multiracial societies breed ethnic division and intractable conflicts by their very nature (as is demonstrated by these very proceedings.) More specifically, in an interview with the Sydney Morning Herald, published on 16 July 2005, I observed that migrants from sub-Saharan Africa were not the only sources of ethnic friction in contemporary Australia. I stood by the comments I had made in a private e-mail to David Shoebridge that East Asian immigration posed other, no less intractable problems, suggesting that one need only “Look at the annual HSC results—the consequence of which is that Oz is creating a new heavily Asian managerial-professional ruling class that will feel no hesitation…in promoting the narrow interests of their co-ethnics at the expense of white Australians.”


Did My Public Comments Breach s18C(1)(a) of the Act?

(i) In relation to black Africans

Mr Hareer alleges not just that my public comments in relation to sub-Saharan African immigration were likely to “offend, insult, humiliate or intimidate” refugees and migrants from Sudan and other African countries but that they “incited hatred” against such persons leading him and his family to feel “great fear of violence and exclusion from the mainstream community.”

In fact, my comments were not likely to offend, insult, or humiliate, much less intimidate a reasonable person of black African ancestry. Mr Hareer is quite simply wrong to assert that I, at any time, alleged “that black people are inferior human specimens and are untrustworthy and criminal by nature.” What I did say is that races differ across a whole range of attributes, carefully noting that no one race can claim comprehensive superiority on every measure of human excellence or fitness. This can be confirmed quite readily by examining the article I have included in the appendix to this response, entitled “Rethinking the White Australia Policy” (due for publication in the Deakin Law Review until political pressure and legal threats from Mr Hareer’s lawyer caused the Vice-Chancellor of Deakin University to order the law review’s editor to pull the article from the current issue), the full, unedited film of interviews with Channels 7 and 9 and also in the transcripts and film footage of the two-hour talk-back program I did with Kwame Koramoah on the African program at Radio Skid Row in Marrickville (the first hour of which was filmed by Channel 7.)

Modern science has confirmed that race is not merely a “social construct.” Despite the myth of equality, race is not merely skin deep. There are significant differences between racial groups in cognitive and athletic ability, behaviour and temperament. Now, as the references in “Rethinking the White Australia Policy” make clear, there are substantial differences between the average IQ levels of sub-Saharan Africans on the one hand and white Europeans on the other with Africans at the lower end of the scale. But, I have stressed also that, if one values athletic and certain sorts of musical ability, for example, black Africans clearly have an edge over both Europeans and East Asians.

Similarly, when it comes to the issue of crime, Mr Hareer clearly does not understand my point. In every interview, I have emphasized that violent criminals, whatever their race, tend to be people with low IQ and poor impulse control. It follows that, if persons of low IQ and poor impulse control are over-represented within a particular racial group, no one should be surprised if that group, as a whole, is over-represented, as well, among people engaging in violent criminal behaviour.

Such observations are not “reasonably likely” to offend everyone of black African ancestry. I have attached an article by Leighton Levy, a black columnist for the Jamaica Star newspaper who has as much claim as Mr Hareer to represent the reasonable black person. In that article discussing “the dark side of black people,” as revealed by the recent Hurricane Katrina disaster in New Orleans, Mr Levy declares that he is “beginning to believe that black people, no matter where in the world they are, are cursed with a genetic predisposition to steal, murder, and create mayhem.” His view is very similar to the point I made in my letter to the Parramatta Sun. If a reasonable black man can express that view, a similar opinion found in my letter to the Sun cannot be characterised as “plainly malicious or scurrilous” and “designed to foster hatred or antipathy” simply because the author happens to be a white man. Discussions about the nature and significance of race must be an equal opportunity activity.

After all, leaders of the Sudanese community in Parramatta themselves have acknowledged, frankly and publicly, the reality of racial differences. So long as they draw attention to the ways in which black Africans shine by comparison with white Australians, African spokesmen, in fairness, must swallow the bitter with the sweet. On 24 August 2005, L’amahz Baz, president of the African Communities Council, was quoted in the Parramatta Sun as follows: “African Americans have dominated athletics and runners from Kenya and Morocco excel in events like the marathon.” Mr Baz clearly is not reluctant to boast of African superiority in sports; indeed, he suggests that Australia will win many more Olympic medals in the future due to the presence of African athletes in this country. Surely, few white Australians feel insulted or offended by his claims of African athletic superiority. Why, then, should Mr Hareer take offence when I, or anyone else, points to areas in which the performance of Africans has been less than stellar, especially in comparison to white Europeans or East Asians (specifically in their comparative cognitive ability or predispositions to the sort of behaviour that is criminalized in almost every society?)


(ii) In relation to Chinese migrants

Mr Hwang and Mr Wong contend that my comments are reasonably likely to offend people of Chinese origin and are capable of inciting racial hatred against them. Such an allegation is demonstrably false; it most definitely flies in the face of my own experience with Chinese students and friends. Certainly, it is hardly credible to suggest that Chinese migrants would take offence at my flattering observation that, on average, their IQ is higher than either black Africans or white Europeans. Nor would any and all Chinese persons automatically take offence at my suggestion that large-scale immigration of East Asians will tend to create serious conflicts of interest between a relatively ethnocentric, Chinese cognitive elite and the predominantly Anglo-European host society. In fact, many Chinese people of my acquaintance have not taken any offence at my public comments on Asian immigration.

I have had many Chinese students and Chinese friends to whom I have on many occasions made such arguments. None of my Asian students have supported the view that my public comments amount to racial vilification. On the contrary, many Chinese and Vietnamese students have stated publicly that I am “not a racist” and that they have never been insulted or offended by the manner in which I have raised these issues in class. Several of my long-time Chinese or Taiwanese friends are writing you to confirm that we have often discussed my views on this subject and, while they may not agree with me, they have never had any reason to feel insulted, offended or humiliated, much less intimidated by such conversations.

I also attach the report of an interview conducted by the German newsmagazine, Der Spiegel, with Lee Kwan Yew, the former Prime Minister of Singapore. In that interview, Mr Lee confirms the essence of the point that I was making in my own interview with the Sydney Morning Herald, namely that in a multiracial society, people act to promote the interests of their own racial and religious groups, even if those ethnic interests might conflict with other economic or social interests. This is the view of a leading Chinese politician, confirming that if Singapore had a democratic polity, “Malays would vote for Muslims, Indians would vote for Indians, Chinese would vote for Chinese.” Clearly, Mr Lee would not be offended by the suggestion that a heavily Chinese ruling elite, once concentrated in Australia’s largest cities, would tend to favour the interests of co-ethnics. That view is also supported by Amy Chua, the Chinese author of an important book (World on Fire: How Exporting Free-Market Democracy Breeds Ethnic Hatred and Global Instability) about “market-dominant minorities” such as the overseas Chinese (I have attached an appreciative review of that book for your information.)

(iii) In relation to people of Mr Wajnryb’s background

There is little that I can say in relation to Mr Wajnryb’s allegation that my comments are likely to offend people of his or his wife’s background. He states that he is the grandson of an immigrant but neglects to specify his grandfather’s race or ethnicity. I assume he is neither of sub-Saharan African nor of Chinese ancestry so I cannot imagine why he should feel offended on behalf of such people. Similarly, in relation to his wife, who is of Argentinian background, it is impossible to know what that means in ethnic terms; she could be of German, Italian or Spanish background for all I know and, if so, would have no reason at all to be offended by anything I have said, no matter how much she or her husband may disagree with my comments.

Certainly nothing I have said could engender “hatred against immigrants to Australia from a non-Anglo-Saxon background.” My argument is simply that a sensible immigration policy, in any nation, would favour the kith and kin of the existing population so as to preserve a desirable ethnic homogeneity. That does not mean that I “hate” or encourage anyone else to hate non-European migrants; it simply means that I believe Australia was a more cohesive society when its people were overwhelmingly of British stock—a view that Mr Wajnryb’s grandfather may well have shared. Following the same logic, Japan should be preserved for the Japanese, just as I believe that Icelanders would be entitled to resent and resist a massive influx of aliens into Iceland. The massive majorities of Icelanders and Japanese who share that view cannot be said, ipso facto, to “hate” foreigners.

There can be little doubt that Japanese (or Chinese, for that matter) patriots would hate to see Japan (or China) become a country like the Sudan, forever fated to be riven by insoluble racial and religious conflicts. There can be no Sudan for the Sudanese simply because there is no agreement among Arab Muslims, black Christians and the welter of other tribal and clan groupings on how to constitute a Sudanese nation. Political unity in polyethnic societies, where it can be achieved at all, is an artificial construct, usually depending upon the creation and maintenance of complex ethnic hierarchies.

Adding the refugees generated by countless black African tribal conflicts and civil wars to the vast numbers of other Third World migrants streaming into Australia will fracture further this country’s already visibly weakening sense of shared national identity. But, patriots who want to preserve Australia for the Australians are not consumed by hatred of African or Asian migrants; instead, they are expressing a deep love for their own people and their own country. No true patriot anywhere, least of all in black Africa or China, will suffer in silence while his beloved ancestral homeland, with its distinctive way of life, is overrun by ever-swelling masses of alien colonists, however peaceful and productive, polite or personally likeable those foreigners may be.

I recognise, of course, that s18C sets a very low threshold; anyone can claim to have been insulted, offended, humiliated or intimidated by my public comments. But I have shown that there are many reasonable people of African or Chinese ancestry who not only would not be offended about my comments on their co-ethnics, they would, in fact, agree with those remarks. But the complainants here may not be such reasonable and open-minded people and may wish to persist in their thin-skinned sense of grievance, transforming mere disagreement into claims of unlawful racial vilification. In that case, the likely effect of my conduct should not be assessed from the perspective of a “hypersensitive victim.” It would be a mistake for the Commission to take the view that insult, offence, humiliation or intimidation is in the eye of the beholder.

The Applicability of s18D Exemptions

Even if the Commission mistakenly adopted a purely subjective test of “insult, offence, humiliation or intimidation,” my conduct could not be characterised as unlawful racial vilification. In what follows, I will demonstrate that my public comments fall squarely within the exemptions to s18C as set out in s18D(b) and (c) of the Racial Discrimination Act.

(i) “Reasonably and in Good Faith”

It is easily demonstrable that any and all of the acts that are the subject of complaints to the Commission were done by me “reasonably and in good faith.” The original letter to the Parramatta Sun, for example, was published only after the letters editor, Mr Gerard Sutton, interviewed me on the phone for upwards of half an hour. During that interview, I explained my reasons for writing the letter and the basis for the argument made therein. Only after Mr Sutton was convinced that I was offering in good faith a reasonable view on the subject of race and immigration did his editor decide to publish my letter along with the front page story and an editorial.

After the letter was published, I received critical e-mail from a number of persons, including Mr David Shoebridge, ( a Woollahra councillor) and Mr Daniel Davidson. To show that my letter was based upon a reasonable understanding of the issues and that I was acting in good faith in criticizing current refugee policy, I responded to my critics with a detailed defence of my position, including in most cases full copies of major sources and references supporting that case by e-mail attachments. My e-mail correspondence with Mr Shoebridge, in particular, was incorporated into Tim Dick’s report in the Sydney Morning Herald on 16 July.

That report, in turn, sparked interest among the producers of the Channel 7 and Channel 9 current affairs programs. In dealing with the producers and journalists from those programs, I, once again, had to demonstrate that I had reasonable grounds for my criticisms of current immigration and refugee policy. I believe I proved my bona fides in the course of my dealings with both Channels. In particular, in putting together Ray Martin’s program for the evening of 19 July, I was confronted on camera for over an hour by a group of about a dozen angry, hostile Sudanese men and women. Anyone who watches the film of that lengthy confrontation will see that I responded in a calm, reasonable manner to a barrage of insults and offensive remarks, trying, as best I could, to demonstrate that I bore the Sudanese no ill will and was concerned only to make a good faith effort to defend the interests of ordinary white Australians.

My good faith willingness to deal openly and honestly with hostile criticism from anyone who claimed to be insulted or offended by my public comments was evident as well in my two-hour appearance on Mr Koramoah’s African community program on Radio Skid Row. Once again, the host and his callers seemed largely uninterested in a reasonable discussion or debate on the merits of ethnic homogeneity or the problems of black crime, in particular. Instead, they happily harangued and insulted me, accusing me of crimes and misdemeanours ranging from my alleged efforts to promote hatred and terrorism in the community to the falsification of my academic credentials. Throughout, I remained calm, taking every opportunity to respond reasonably and in good faith, that is to say, without malice, to every point raised by every speaker.

My determination to act reasonably and in good faith, whatever the provocations aimed at me, was also on display at the “Racism Within” forum held on 5 August at Macquarie University. This event was sponsored by several academic departments of the University and had the blessing of the Vice-Chancellor’s office. The meeting was organized around a panel discussion between seven academic staff members of the threat posed to pluralism and our “common humanity” by my “hateful” and “racist” public comments. Members of the three- to four-hundred person strong audience were also invited to speak. It was made clear to me before the meeting began that I would not be permitted to participate as a member of the panel. I could attend as a member of the audience but would be allowed only two minutes to speak.

The forum was essentially a show trial in which the panel members and audience joined in ritual denunciations of my “racism,” many calling for my full public confession of wrong-doing. One member of the audience labelled me a “racist scumbag” with the apparent approval of the meeting’s chairman. I accepted this lengthy excoriation with as much dignity as I could muster and, in the two minutes allotted to me at the end of the meeting, I sought to make a brief case for the reality of racial differences. Not surprisingly, in what has taken on the atmosphere of a lynch mob, I was howled down and soon surrounded by angry Africans. Even then, I did my best to respond reasonably and in good faith to their heated comments until concerned security officers hustled me out the back entrance.

My conduct will stand the most rigorous, objective test of reasonableness and good faith behaviour. One final example was provided by the anonymous referees assessing the article “Rethinking the White Australia Policy,” that I had been invited to submit to the Deakin Law Review. They read my initial offering, made criticisms, comments and suggestions for change. In light of their initial remarks, I made substantial revisions to the draft article and re-submitted the piece. On reading the revised article (which explicitly articulated and expanded upon the arguments I had made publicly in relation to Asian and African immigration), the referees declared that they were satisfied that the article presented a reasonable, well-constructed argument that deserved to be published in a scholarly journal. And so it would have been, but for the intimidating tactics employed by Mr Hareer’s lawyer against Deakin University to stop its publication.

It would be impossible for the Commission to make a finding that my conduct throughout this entire affair has in any way smacked of “dishonesty or fraud; in other words [of] something approaching a deliberate intent to mislead or…a culpably reckless and callous indifference” to the possibility that Africans or Chinese persons might take offence at my public comments. I have done everything reasonably possible in the circumstances to assuage the feelings and meet the concerns of my critics of whatever race.

(ii) Genuine Academic Purpose

In this matter, it has been my opponents who have acted unreasonably and in bad faith. By contrast, my acts have been good faith and reasonable efforts to carry on activities serving genuine academic purposes while also making fair comments, based on my genuine belief, on matters of public interest. In what follows, I will provide evidence in support of both aspects of that proposition.

The complaints here relate to things that I have said or done, “reasonably and in good faith,” as we have just seen, “in the course of” statements, publications, discussions or debates” that were made or held for genuine academic purposes. As is well known, academics are expected to engage in various forms of community outreach apart from their normal teaching responsibilities and scholarly research. Community outreach is usually related in some way to an academic’s area of expertise. In my case, it was my teaching and research experience in the areas of American constitutional history (and the associated problems of race relations) and Australian immigration law and policy that gave me special knowledge of issues related to race, racial differences and immigration. Accordingly, I was able to offer the readers of the Parramatta Sun a unique academic perspective on the question of Sudanese immigration into our community.

Indeed, it was because I am an Associate Professor of Public Law that Gerard Sutton, the letters editor, decided to take my letter seriously. Soon after receiving that letter, Mr Sutton had a long phone conversation with me in which I explained that, in my view, the academic world is on the cusp of a paradigm shift in thinking about race and racial differences. For the past half century or so, academics across many disciplines have accepted the egalitarian view that racial differences either do not exist or are merely superficial. But, as James Mackintosh makes clear in his letter to the Commission, the work of psychologists such as Arthur Jensen and J Phillipe Rushton (from whom you may also expect a letter), paleo-anthropologists such as Vincent Sarich and numerous geneticists and medical scientists has rendered the egalitarian myth scientifically unsustainable.

The work of scientists such as Rushton and Jensen was highly controversial ten or fifteen years ago; they were rogue academics working in isolation. But, as is clear from the Ottawa Citizen article by Alan Duffy, “Revisiting Rushton,” (included in the appendices attached herewith), today Rushton is part of “a small army of scientists…exploring the genetic foundation of intelligence, and the genetic differences between people of African, Asian, Indian, Middle Eastern and European descent.”

My public comments on IQ differences between racial groups and the behavioural and temperamental differences that also help to explain comparative propensities toward criminal behaviour drew upon my reading of the work of people such as Rushton and Jensen, Michael Levin and Vincent Sarich and Frank Miele. In my own research and scholarly publications over the past few years, I have made my own contribution to the academic debate around issues of race, multiculturalism and mass Third World immigration into the Western world. As I made clear to every journalist from the print and electronic media to whom I spoke, my public comments grew out of both the shifting paradigm of academic discourse on race, in general, and my own work which attempts to explain the implications of the new racial science for a wide range of public policy issues.

That fact was recognized most recently by the anonymous referees for the Deakin Law Review who approved the publication of an article, “Rethinking the White Australia Policy,” repeating at greater length the arguments and analysis that I tried (with only limited success) to squeeze through the media filtering process over the past few months. Their assessments provide an objective foundation for the conclusion that the acts or statements which have become the subject of complaints to the Commission were said or done as part of a continuing course of statements, publications, discussions and debates which have been made or held for genuine academic purposes. Further support for that conclusion will be found in letters to the Commission from James Mackintosh, Frank Salter and J Phillipe Rushton. Everything I have said or done over the past few months bearing on the subjects of race and immigration can also be justified as “fair comment” on matters of public interest made as an expression of genuine belief on my part. It is to that issue that I now turn.

(iii) Fair Comment

Even if I had not been an academic, my public comments would not amount to unlawful racial vilification because they are nothing more than fair comment, based upon a genuine belief on my part, on matters of public interest. The mere fact that so many different media organizations in so many states and even other countries for so long found my public comments significant and worthy of extended discussion and debate demonstrates, beyond question, that they touched on matters of public interest. That my comments on the downside of both Asian and African immigration were fair and within the legitimate scope of public debate is confirmed by a number of things. One could point to the overwhelming response in the Channel 9 phone poll asking viewers of Ray Martin’s A Current Affair whether they agreed with my view that Asian and African immigration were bad for Australia. Over 36,000 callers responded with 85% expressing agreement with my position.

It must also be significant that the Higher Education Supplement in The Australian chose to publish an edited version of my article on “Rethinking the White Australia Policy.” So, too, did the OnlineOpinion.com.au. That latter version included the passages from that article dealing with racial differences in cognitive ability as well as the discussions of black criminality and the potential problems posed by “market-dominant minorities” such as the overseas Chinese. Clearly, the editors of OnlineOpinion were right in their assessment that my remarks on such subjects are fair comment on important matters of public interest. There has been an overwhelming reader response, with over 120 comments at the time of writing.

Many of those responding to my work will, of course, disagree with it. But few agree that my views should be forbidden public expression. Even many of those who disagree with my views will have to concede that considerable evidence can be advanced in support of my position. To demonstrate that my views rest upon a solid evidentiary foundation is perhaps the best way to establish that they represent fair comment on what no one could deny are matters of public interest.

The public comments that have been the subject of complaint to the Commission fall into two categories: those relating to black African migrants and those dealing with the problems posed by East and South Asian immigration. On the first issue, the complaints concentrate on my suggestion that “experience practically everywhere demonstrates that a large and expanding black population is a sure-fire recipe for increases in crime, violence and a wide range of other social problems.” Complaints relating to the second issue cite my prediction that large-scale East Asian immigration will produce a heavily Asian ruling class that will favour the interests of co-ethnics at the expense of white Australians. To show that my public comments on both issues represent fair comment on matters of public interest, I will review the evidence supporting my observations and predictions, beginning with the issue of black criminality. I will then address the problem of “market-dominant minorities” posed by continued large-scale immigration from East Asia, in particular.

(a) The problem of black criminality

As suggested above, the problem of violent crime, within any racial group, is associated with people characterised by low IQs and poor impulse control. The evidence that persons of black African ancestry are, on average, more likely than white Europeans or East Asians to have low IQs is overwhelming. It is also clear that black Africans have higher levels of the serum testosterone associated with poor impulse control. Not surprisingly, therefore, black Africans are, by comparison with white Europeans and, especially, East Asians, much more likely to be involved in violent street crime. (I have identified the major sources providing support for that claim in footnotes 8 and 52-54 in “Rethinking the White Australia Policy,” included in the appendix.)

Apart from the academic sources cited in “Rethinking,” there exists a plethora of readily available journalistic evidence demonstrating that black criminality is a widely recognized problem not just overseas but already in Australia. In the appendix, I have provided copies of newspaper articles detailing the upsurge of crime in Dandenong, Victoria that has been associated with the recent influx of Sudanese refugees. The outbreak of massive civil disorder in the wake of Hurricane Katrina sparked widespread discussion of the taboo topic of black criminality in the USA. I have included several pieces by Steve Sailer, a well-known commentator on racial issues, which make it clear that blacks are, indeed, more prone to violent crime than other racial groups.

That conclusion has been reinforced by the recent release of a report compiled by the New Century Foundation entitled The Color of Crime. The appendices include a piece by Jared Taylor pointing out, inter alia, that, according to US government statistics, blacks in the USA “are just 13 percent of the population but they commit more than half the muggings and murders in the country.” In general, the best single indicator of how dangerous an area is turns out to be the proportion of blacks and Hispanics in the population. Moreover, blacks commit more violent crime against whites than against fellow blacks or Hispanics. As the piece by John Woods (also included in the appendices) indicates much the same picture emerges from Home Office crime statistics in the United Kingdom.

My public comments on the potential for the increases in crime and violence that might be associated with immigration from the black populations of sub-Saharan Africa were solidly grounded in a study of the information available through a wide variety of academic, official and media sources. I accept the credibility of those sources, and my own comments were a fair summary of that material. The same conclusion applies to my comments on the very different dangers posed by Asian immigration, particularly from northeast Asia, i.e. China, Korea, Japan and Taiwan.

(b) East Asian cognitive elites and the problem of “market-dominant minorities”

Mr Hwang and Mr Wong object to my statement that, if present trends continue, in twenty or thirty years time Australia will have a “market-dominant minority” of overseas Chinese likely to favour the interests of co-ethnics at the expense of white Australians. My prediction is based upon Philippines-born, Chinese-American, Amy Chua’s analysis (in her book, World on Fire) of the intractable conflicts between the overseas Chinese and the host populations of just about every country in Southeast Asia. I also draw on discussions of the overseas Chinese to be found in the works cited in footnotes 44 and 47 in “Rethinking the White Australia Policy.” In the Philippines for example, the Chinese minority, representing only 1 percent of the population, controls over 60 percent of the economy. Such a lopsided situation generates widespread resentment among Filipinos. This simmering tension boiled over onto Chua’s own family when an elderly aunt was murdered by her Filipino chauffeur, a crime all but ignored by Filipino police. Similar conflicts are an inescapable fact of life almost everywhere in Southeast Asia, most famously in Malaysia where the dominance of the overseas Chinese forced native Malays to adopt various forms of “affirmative action” to protect their ethnic interests.

In his review (included in the appendices) of Chua’s book, Matt Nuenke suggests that the best explanation for the ability of the Chinese to establish and maintain their position as a “market-dominant minority” is the significant IQ gap (in favour of the Chinese) between them and the native populations. The IQ differential between Chinese and white Australians is not as large but it does exist and already it has had a striking effect on the competition for places in higher education and access to professional careers—with white Australians being the big losers.

There are also good reasons to expect that a market-dominant minority of East Asians would tend to practice forms of ethnic nepotism usually frowned upon by more individualistic white Australians. This is certainly the norm for both Chinese both overseas and at home. As discussed above, Lee Kwan Yew simply takes it for granted that Chinese in a multi-racial society will always favour their co-ethnics. It is well known that Chinese have long harboured deeply xenophobic, even downright “racist” attitudes. Such ethnocentric attitudes have been powerfully reinforced in recent times by a Communist government that, having lost its Marxist ideological mooring, is fearful of losing control over its vast empire. Immigrants from mainland China have been taught to hate “the foreign devils” and cherish the Motherland, “which never has done, and never could do, any wrong.” Steeped from childhood in an ever-more aggressive Chinese nationalism, such immigrants are unlikely to resist powerfully ingrained habits of ethnic nepotism. Indeed, John Derbyshire (in an article also included in the appendix) warns that Chinese immigrants pose the very real danger of an imported Sino-Fascism.

That danger may become ever more pronounced as China itself advances in military and economic might. But, even if we leave that geo-political dimension of the problem aside, there is no denying that individualistic white Australians, taught from infancy that white racial pride is a grave moral failing and that ethnic nepotism is an unlawful form of racial discrimination, will be extraordinarily vulnerable to competition from a much more cohesive cognitive elite of overseas Chinese. Janet Landa points out that in overseas Chinese society, Confucian ethics prescribe “differences in patterns of mutual aid obligations between people with varying degrees of social distance within a well-defined social structure-near kinsmen (e.g. family members), distant kinsmen in extended family and lineage, clansmen, fellow villagers, and people speaking the same dialect.”

The strongest ties are within the family where social distance is at a minimum. Trustworthiness in trade relations is generally measured in terms of concentric circles extending outward from family and near kin. On Landa’s analysis, “Chinese social structure, unlike Western social structure which is individualistic in nature, consists of a careful ranking of people who are classified according to distinct categories of social relationships.” Clearly, the greater in-group solidarity of Chinese operating within such a social structure gives them a powerful edge in competition with unorganised, individualistic white Australians. For white Australians, even the nuclear family had lost much of its former power to bind people together in cohesive units.

Whether one agrees with that analysis or not, it is impossible to deny that any public comments based upon that evidentiary foundation represent “fair comment” within the meaning of s 18D(c)(ii) of the Racial Discrimination Act. It follows that none of the statements or acts that are now the subject of complaints to the Commission can be considered unlawful racial vilification.


Conclusion

In light of the foregoing analysis, I submit that all of the four complaints ought to be summarily dismissed as altogether lacking in substance. Indeed, it is clear that they are best characterized as frivolous and vexatious in nature, being designed mainly to end public discussion of important matters of public interest.

NOTE: Since the above was written, the two complaints submitted by Chinese have been dismissed on the grounds that the remarks attributed to Prof. Fraser by the SMH were originally made in a private e-mail. Another complaint, from the guy married to the Argentinian, was dismissed on the grounds that he was of some indeterminate ethnicity uninvolved in any of Prof. Fraser's public comments. As a consequence, the only live complaint at the present is the one submitted by Newhouse Lawyers, ostensibly on behalf of Mr Hareer, a Sudanese colony leader.

Tuesday, March 21, 2006

SEX AND IQ

The ferocity of the attack on the recent Lynn & Irwing paper is exactly what we would expect of a genuine contribution to the advancement of knowledge. Work that overturns what has long been believed will of course be resisted. To allow the general reader to see for himself/herself that the paper is utterly sound, I have reproduced it below. Out of respect for copyright, however, I have omitted the Tables and References (Essential to serious scholars)

Citation:
Lynn, R. and Irwing, P. (2004) Sex differences on the Progressive Matrices: a meta-analysis. Intelligence, 32, 481-498.

Sex differences on the Progressive Matrices: a meta-analysis

Richard Lynn [a], Paul Irwing [b][*]

[a] University of Ulster, Coleraine, Northern Ireland
[b] University of Manchester Institute of Science & Technology, Manchester, M60 1QD, UK
[*]Corresponding author. E-mail address: Paul.Irwing@umist.ac.uk

Abstract

A meta-analysis is presented of 57 studies of sex differences in general population samples on the Standard and Advanced Progressive Matrices Results showed that there is no difference among children aged 6 to 14 years, but that males obtain higher means from the age of 15 through to old age. Among adults the male advantage is .33d equivalent to 5 IQ points. These results disconfirm the frequent assertion than there are no sex differences on the Progressive Matrices and support a developmental theory that a male advantage appears from the age of 15 years. A meta-analysis of 15 studies of child samples on the Colored Progressive Matrices showed that among children aged 5-11 years boys have an advantage of .21d equivalent to 3.2 IQ points.

1. Introduction

During the last two decades the issue of sex differences in cognitive abilities has been addressed by carrying out meta-analyses of studies on verbal abilities (Hyde and Linn, 1988), spatial abilities (Linn and Peterson, 1985; Voyer et al., 1995) and mathematical abilities (Hyde et al., 1990). No meta-analysis has yet been made of sex differences in reasoning ability. In this paper we present the first study of this issue in the form of a meta-analysis of sex differences in non-verbal reasoning ability measured by the Progressive Matrices Test.

Raven's Progressive Matrices was constructed in the late 1930s as a test of non-verbal or abstract reasoning ability (Raven, 1939). It has become one of the leading and most frequently used tests of this ability and has been described as "the paradigm test of non-verbal, abstract reasoning ability" (Mackintosh, 1996 p. 564). It is also widely regarded as the best or one of the best tests of Spearman's g, the general factor underlying all cognitive abilities. Thus, Court (1983, p. 54) has written that it is "recognised as perhaps the best measure of g"; and Jensen (1998, p.541) that "the Raven tests, compared with many others, have the highest g loading". There are three versions of the test: the Standard Progressive Matrices for the ages of 6 years to adulthood; the Colored Progressive Matrices, an easier version of the test designed for children aged 5 through 12; and the Advanced Progressive Matrices, a harder version of the test designed for older adolescents and adults with higher ability.

The issue of whether there are any sex differences on the Progressive Matrices has frequently been discussed. It has been virtually universally concluded that there is no difference in the mean scores obtained by males and females. The first statement of this conclusion was made by Raven (1939,p.30) who wrote that in the standardisation sample "there was no sex difference, either in the mean scores or the variance of scores, between boys and girls up to the age of 14 years. There were insufficient data to investigate sex differences in ability above the age of 14". The conclusion that there is no sex differences on the Progressive Matrices has been endorsed by numerous scholars. For instance, Eysenck (1981, p.41) stated that the tests "give equal scores to boys and girls, men and women". Jensen (1998, p.541) writes "there is no consistent difference on the Raven's Standard Progressive Matrices (for adults) or on the Coloured Progressive Matrices (for children). Mackintosh (1996, p.564) writes "large scale studies of Raven's tests have yielded all possible outcomes, male superiority, female superiority and no difference". From this he concludes that there is no difference or, in a subsequent paper, that there is only a very small difference consisting of no more than a difference of 1 to 2 IQ points among adults "either way" (Mackintosh, 1998).

Both Jensen and Mackintosh reply for their conclusion that there is no sex difference in mean scores on the Progressive Matrices on a review by Court (1983). This review summarized 118 studies of sex differences on the Progressive Matrices and concluded that most showed no significant difference in mean scores although some showed higher mean scores for males and others found higher mean scores for females. From this he concluded that "there is no consistent difference in favour of either sex over all populations tested.the most common finding is of no sex difference. Reports which suggest otherwise can be shown to have elements of bias in sampling" (p.62) and that "the accumulated evidence at all ability levels indicates that a biological sex difference cannot be demonstrated for performance on the Raven's Progressive Matrices" (p.68).

Court's review is a useful starting point for tackling the question of sex differences on the Progressive Matrices but it cannot be accepted as satisfactory basis for the conclusion that no sex differences exist. The review has at least five deficiencies. First, it is over twenty years old and a number of studies of this question have appeared subsequently and need to be considered. Second, it is a literary review that does not attempt to quantify the magnitude of any sex difference that may exist. Third, it includes general population samples and a variety of convenience samples including those of psychiatric patients, deaf children, retarded children, shop assistants, clerical workers, British, Indian and French university students, Native Americans and Inuit. Many of these samples cannot be regarded as representative of males and females and in some cases this is stated by the authors and reported by Court. For instance, two studies showing that among British military personnel in World War 2 female neurotic patients obtained higher scores on the Progressive Matrices than males; Court reports that the authors of the studies believed that the reason for this was "probably due to the biased selection procedures in favour of bright women in the British Armed Forces" (Court, 1983, p.60).

With such diverse and unrepresentative samples in Court's review, it is not surprising that in some studies higher scores were achieved by males, in others higher scores were achieved by females and in others there were no sex differences. Fourth, Court does not provide information on the sample sizes for approximately half of the studies he lists, and where information on sample sizes is given the numbers are generally too small to give a statistically significant difference between males and females. To detect a statistically significant difference of between 2 to 5 IQ points such as may exist between males and females on the Progressive Matrices requires a sample size of around 500. Court's review gives only one study of adults with a sample size of this number or more. This is Heron and Chown's (1967) study (n= 600) on which men obtained a significantly higher mean score than women of 0.31d, approximately 4.65 IQ points (see Table 1). Nine of the studies showing no statistically significant sex differences in Court's review have sample sizes of fewer that 100, such as those of 60 Welsh 11-12 year olds and 22 American 5 year olds. Studies with these small sample sizes that show no significant differences between males and females have no value for the resolution of the issue of whether there is a small but significant sex differences on the Progressive Matrices. We found that of the 118 studies covered in Court's review only 21 met the criteria of being general population samples and having sample sizes on a minimum of 50 males and 50 females that we have adopted in the meta-analysis to be presented. Fifth, Court's review did not include all the studies; in our literature search we found nine studies of general population samples that were not given in the review and these included two with largest sample sizes (Hsu, 1976; Wilson, De Fries, McClearn, Vandenberg, Johnson & Rashad,1975). For all these reasons, Court's review cannot be accepted as an adequate basis for the conclusion that there are no sex differences on the Progressive Matrices.

The consensus that there are no sex differences on the Progressive Matrices and on other tests of abstract (non-verbal) reasoning ability has been challenged by Lynn (1994, 1998, 1999). He has proposed a developmental theory of sex differences in intelligence that states that boys and girls mature at different rates both physically and mentally during childhood and adolescence. Boys and girls mature at about the same rate up to the age of around 7 years; from the age of 8 girls begin a growth spurt in which there is an acceleration of their physical growth in respect of height, weight and brain size; the growth rate of girls slows at the age of 14 and 15, while the growth of boys continues. The developmental theory states that intelligence follows the same trend. Evidence supporting the theory has been provided in Lynn (1994, 1998, 1999) and in Lynn, Allik and Must (2000). In regard to abstract (non-verbal) reasoning ability, the theory as originally formulated in Lynn (1994) states that over the age range of around 9 through 12 years girls have an advantage of approximately 1 IQ point; by the age of 16 years this has changed to a small advantage in favor of boys and among adults the male advantage is 2.4 IQ points. These estimates were not derived from data on the Progressive Matrices but, in the case of adults, from the American standardization samples of the Differential Aptitude Test. In a subsequent compilation of studies it was proposed that among adults the male advantage on abstract reasoning is approximately 5 IQ points (Lynn, 1999).

In addition to the theory that sex differences on the Progressive Matrices vary by age, there is also a theory proposed by Mackintosh (1998) that there are cohort differences such that among older generations men achieved higher means than women but this is no longer the case among recent cohorts of young adults.

The resolution of these conflicting theories requires a meta-analysis of studies on sex differences on the Progressive Matrices such as we have carried out and now describe. The meta-analysis is designed to test three hypotheses. These are:

1. The Eysenck (1981)-Court (1983) -Mackintosh (1996) -Jensen (1998) hypothesis that there is no sex difference on the Progressive Matrices.

2. The Lynn (1994, 1998, 1999) hypothesis stating that there is no sex difference on the Progressive Matrices among young children up to the age of 8 years; that girls have a slight advantage from the ages of 9 through 12 years; that boys and girls obtain about the same mean scores at the ages of 13-15 years; that at the age of 16 years boys have a higher mean than girls; that this advantage increases up to the age of 18-19 years; and that from the age of 20 onwards the male advantage lies between 2.4 IQ points (Lynn, 1994) and 5.0 IQ points (Lynn, 1999).

3. The Mackintosh (1998, p. 538) hypothesis that the higher mean scores of men on the Progressive Matrices is a cohort effect such that it may have been present among older generations but that "the sex difference in general intelligence among young adults today in the USA, Britain or Israel" (and, presumably, elsewhere in the economically developed world) " is trivially small, surely amounting to no more than 1-2 IQ points either way", i.e. in favor of either males or females.

2. Method

Criteria for selection of studies

The meta-analyst has to address three problems before analysing the data. These have been identified by Sharpe (1997) as the "Apples and Oranges", "File Drawer" and "Garbage in - Garbage out" problems. The "Apples and Oranges" problem is that different phenomena are sometimes aggregated and averaged, where aggregation shows different effects for different phenomena. For instance, Hyde and Linn (1988) have shown in a meta-analysis of sex differences in verbal abilities that the magnitude of the difference varies for different kinds of verbal ability and ranges from a female advantage of .33 d for speech production to a male advantage of .16 d for analogies. This indicates that verbal ability is too broad a concept for the calculation of sex differences and that it is more meaningful to calculate differences in narrower abilities such as essay writing, vocabulary, verbal reasoning, etc. The best way of dealing with this problem is to carry out meta-analyses in the first instance on narrow abilities and then attempt to integrate these into broader categories. In the present meta-analysis, this problem has been dealt with by confining the analysis to studies using the Progressive Matrices.

The "File Drawer" problem is that studies producing significant effects tend to be published, while those producing non-significant effects tend not to be published and remain hidden in the file drawer. This is a serious problem for meta-analyses comparing effects of treatments, such as whether various methods of psychotherapy have any beneficial effect, in which studies finding positive effects are more likely to be published, while studies showing no effects are more likely to remain unpublished in the file drawer. It is considered that this should not be a problem for our present inquiry because very few studies have ever been carried out with the primary objective of ascertaining whether there are sex differences on the Progressive Matrices. Data on sex differences on the Progressive Matrices are available because they have been reported in a number of studies as a by-product of studies concerned with other phenomena.

The "Garbage in - Garbage out" problem is concerned with what to do with poor quality studies. Meta-analyses that include many poor quality studies have been criticised by Feinstein (1995) as "statistical alchemy" which attempt to turn a lot of poor quality dross data into good quality gold. Poor quality studies are liable to obscure relationships that exist and can be detected by good quality studies. Meta-analysts differ in the extent to which they judge studies to be of such poor quality that they should be excluded from the analysis. Some meta-analysts are "inclusionist" while others are "exclusionist" in the terminology suggested by Kraemer, Gardner, Brooks and Yesavage (1998). The problem of what should be considered "garbage" and therefore excluded is a difficult one for meta-analysts. For our own meta-analysis, poor quality studies are of two general kinds.

First, those whose samples are unrepresentative of males and females, such as those of shop assistants, clerical workers, psychiatric patients in the military and the mentally retarded included in Court's review. To deal with this problem, our meta-analysis is confined to general population samples. Many of the general population studies are standardisation samples that have been selected to be representative of the general population. Others have been less rigorously selected but nevertheless have been obtained in such as way that there is no reason to suspect any bias in the selection of males and females.

The second kind of poor quality study consists of those with small sample sizes that are liable to produce anomalously large chance effect sizes that obscure the true relationship. Some meta-analysts ignore differences in sample sizes and accord all studies equal weight irrespective of sample size. This is reasonable for certain data sets where all the studies have about the same sample sizes. Where this is not the case, some meta-analysts deal with this problem by ignoring studies with samples below a certain size, while others weight the studies by the sample sizes.

These two solutions amount to much the same thing because weighting by sample size dilutes and may effectively eliminate the contribution of studies with small samples. Where the meta-analyst has a number of large samples the simplest procedure is to ignore small samples and confine the analysis to studies where sample sizes are considered acceptable. Data providing sex differences on the Progressive Matrices differ considerably in the sample sizes, ranging from 43,825 for 5 year olds on the Coloured Progressive Matrices in Taiwan (Hsu, 1976) to 22 for 5 year olds on the same test in the United States (Garrity and Donaghue, 1976). To accord these two studies equal weight in a meta-analysis cannot be regarded as a reasonable procedure. The solution adopted in the present meta-analysis has been to use only data based on a minimum of 50 males and 50 females. The inclusion of such small sample sizes is probably vulnerable to the criticism of being too inclusionist. On the other hand, it is considered preferable to err on the side of over-inclusion because this provides other meta-analysts with the references and data that can be reworked by in various ways to reduce the effects of small samples.

Literature Search

Once the meta-analyst has drawn up criteria for studies to be included in the analysis, it is necessary to obtain all the studies meeting these criteria. This is a difficult problem and one that it is rarely possible to solve completely. Meta-analysts attempt to find all the relevant studies of the phenomena being considered by examining previous reviews and searching abstracts. But these do not identify all the relevant studies, a number of which provide data incidental to the main purpose of the study and which are not mentioned in the abstract or among the key words. Hence the presence of these data cannot be identified from abstracts or key word information. Many studies of this kind can only be found by searching through a large number of publications. It is virtually impossible to identify all relevant studies. For the present meta-analysis, the studies were obtained from Court's (1980, 1983) bibliography and review of studies of the sex difference on the Progressive Matrices, from the series of manuals on the Progressive Matrices published by Raven and his colleagues (e.g. Raven, 1981, Raven, Court and Raven, 1996, Raven, Raven and Court, 1998), and from Psychological Abstracts from 1937 (the year the Progressive Matrices was first published). In addition to consulting these bibliographies, which are widely regarded as comprehensive, we conducted computerized data base searches of PsycINFO, ERIC, Web of Science, Dissertation Abstracts, the British Index to Theses and Cambidge Scientific Abstracts for the years covered up to and including 2002. Finally, we contacted active researchers in the field and made a number of serendipitous discoveries in the course of researching this issue. Our review of the literature covers the years 1939 to 2002.

Organization of Meta-analyses

The studies have been organized into two categories. The first consists of studies of general population samples for the Standard and Advanced Progressive Matrices and presents sex differences for individual years for the ages 6 through 19 and for 10 year age groups of 20-29, 30-39, etc. through 80-89. Data for the Standard and Advanced Progressive Matrices are combined because it is considered that the Advanced Progressive Matrices measures the same non-verbal reasoning ability as the Standard Progressive Matrices. There are a few data sets whose numbers are too small to meet the criterion of a minimum of 50 males and 50 females for individual years (i.e. for 5s, 6s etc.) but for which data for several years can be averaged to meet the criterion. In these cases the sex differences for several years are averaged and entered for the mean year (for instance, data for 50 6 year olds, 50 7 year olds and 50 8 year olds would be averaged and entered as 7 year olds). There are a few data sets where the sex difference is given for a sample with an age range but not for individual years (e.g. for 11-13 year olds but not separately for 11, 12 and 13 year olds). In these cases the sex differences are entered for the average year. The second category consists of data for the Coloured Progressive Matrices. The same methodology is used as for the Standard and Advanced Progressive Matrices.

Strategy of Analysis

The analysis followed procedures developed by Hunter and Schmidt (1990). Cohen's d (the difference between the male and female means divided by the pooled standard deviation) was adopted as the measure of effect size (Cohen, 1977). In the majority of studies, means and standard deviations were reported, which allowed direct computation of d. In a minority of cases, estimates of the standard deviation were obtained from tables of percentiles. Otherwise, in a very few cases, the effect size was derived from a t ratio using the conversion formula provided by Rosenthal (1991).

Meta-analysis of effect sizes. First, the mean of effect sizes was calculated for each age group, weighted by sample size (N-1) (see Table 3). These estimates were then corrected for measurement error. The weighted artefact distributions used in this calculation were derived from those reliability studies reported in Court and Raven (1995) for the Standard, Advanced and Coloured Progressive Matrices, with a sample n 3 300. In order to detect the presence of moderator variables, tests of homogeneity of effect sizes were conducted using Hunter and Schmidt's 75% rule. Each corrected mean d-score was fitted with a confidence interval. We did not compute credibility intervals, since the 95% confidence intervals for corrected mean d-scores were generally wide (see Table 3), which would suggest that the boundaries of credibility intervals might have been in error by some margin.

3. Results

The results of the studies on sex differences on the Standard and the Advanced Progressive Matrices are shown in Table 1. The table gives data derived from 57 studies analysed to provide effect size estimates for 195 samples, with participants numbering a total of 80,928. Samples were considered to be independent within age categories, since no study provided more than one estimate per age group. The table gives the location of the study; the size of the male and female samples; the male-female difference in d-scores with positive signs denoting higher means by males and negative signs higher means by females; the reference; and whether the Standard or Advanced form of the test was used, identified in the notes to the table. The number of effect sizes for each age varied from 1 to 23, and the pooled sample sizes ranged from 200 to 10,708 across age groups. For some age groups (ages 6, 19, 70-79 and 80-89 years old) the sample sizes are small. The reader should, therefore, exercise caution in interpreting some point estimates, and use the width of confidence intervals as an indicator of the accuracy of corrected mean d-scores (see Table 2).

Schmidt and Hunter corrections for unreliability are controversial (e.g. Rosenthal, 1991). For this reason, Table 2 gives uncorrected mean d scores, in addition to the corrected mean d coefficients and their associated confidence intervals. The combined reliabilities and variance of reliabilities for the two types of test (APM and SPM), which are necessary in order to calculate the corrected d coefficients, were computed using standard weighting procedures (Hunter & Schmidt, 1990), for each age group, separately. The estimates for reliability and variance, in this instance, have narrow confidence intervals. This suggests that the corrected d coefficients provide a better estimate of the population difference as compared with the uncorrected coefficients: in any case there is little difference between them.

The general trend of the data on the Standard and Advanced Progressive Matrices is that boys obtain slightly but not significantly higher means over the ages 6 through 9 years; this is followed by slight but not significantly shift towards higher means for girls over the age range from 10 through 13 years. At the age of 14 an advantage of .08d for boys begins to appear, increasing to .10d and becoming statistically significant at the age of 15 and increasing further to a statistically significant .17d at age 19. The male advantage becomes .33d among adults young adults aged 20-29 and remains at approximately this size through all later age groups to 80-89. The sex difference is statistically significant for all the adult age groups except among the 80 to 89 year age group, probably because of the small sample size.

The shape of the growth curve is not established precisely in the current data, but it is clear that whereas there is no sex difference on the Progressive Matrices among younger children, a difference in favor of males appears in adolescents at about 14 to 15 years of age and this difference increases in later adolescence and among adults. As indicated by the tests for homogeneity (see Table 2), the fluctuations in the point estimates among adults, within and across age groups, should probably be attributed to sampling errors. The aggregated mean of the corrected d-scores for the adult age groups is .33d, equivalent to 5 IQ points. In view of the homogeneity test results, this is the best estimate for the male advantage on the Progressive Matrices among adults.

Table 3 gives results for the Colored Progressive Matrices. There are 15 studies which yielded 42 estimates that were independent within age groups, and comprised a total sample with n = 60,168. The study of Hsu (1976) provided a disproportionate number of the 6 year olds. However, inspection of Table 3 shows that the median d (.14) for that age group does not differ markedly from the weighted mean d. It may, therefore, be concluded that this study does not distort the estimate of d for 6 year olds. Examination of the sex differences for each year shows that there was no sex difference at 5 years of age, boys obtain significantly higher means (p< .05) of .21d, .24d and .34d for children aged 6, 7 and 8 years, respectively; there is no difference at ages 9 and 10 years; and boys obtain a significantly higher mean of .30d at age 11 years. (A mean effect size is statistically significant when the confidence intervals do not include zero.) The homogeneity tests (see Table 2) indicate that within age group differences largely reflect sampling error and to a lesser extent variation due to attenuation. Exceptionally, at age 8, artefacts explained only 45% of variance in d-scores, suggesting, somewhat anomalously, the presence of a moderator. Additionally, the homogeneity test for all children (45% of variance explained by artefacts) suggests that differences among age groups reflect both real differences and sampling error. Since it is not possible to disentangle real differences from sampling error, a reasonable treatment of the data is to average the results for the six age groups giving a corrected weighted mean of .21d, equivalent to a higher mean for boys of 3.2 IQ points. The advantage of boys on the Standard Progressive Matrices averaged over the same age range is only .02d. Homogeneity of effect sizes is also supported for children of 6 to 8 years of age, whereas heterogeneity of effect sizes is observed for children and adolescents from 9 to 18 years of age, indicating the presence of one or more moderator variables.

We now examine Mackintosh's (1998) suggestion that while there may have been a sex difference on the Progressive Matrices in earlier times it became virtually zero in the late twentieth century. In addition, we examine a possible explanation of the heterogeneity of effect sizes observed for children and adolescents from 9 to 18 years of age, which may reside in different rates of maturation across ethnic groups. Analyses to test for these possibilities were carried out using maximum likelihood based weighted regression analysis, as implemented within the ANOVA and regression programmes of STATA. Since these procedures produce correct estimates of sums of squares, the corrections prescribed by Hedges and Becker (1986) for standard weighted regression were unnecessary.

In the first analysis, at step 1, age was entered as a control variable. The mid-point of the range served as the estimate of age in adult samples. At step 2, we entered the birth date of the cohort, estimated by subtracting the age of each sample from the date on which the study was published, in order to test for generational effects. As previously established, the analysis revealed a significant effect of age on effect size (DR2 = .19, ? = .44 , t(1,162) = 5.14, p = .00). However, once age was controlled there was no evidence of a significant effect of the birth date of each cohort (DR2 = .00, ? = .01, t(1,162) = .07, p = .95) on the magnitude of the sex difference with respect to the Progressive Matrices. Consequently, Mackintosh's suggestion that sex differences on the Progressive Matrices have declined among the younger generation is not supported.

In the second analysis, at step 1 we again entered age and, at step 2 we entered ethnicity, which was coded (1 = East Asian, 2 = White, 3 = South Asian, 4 = Black, Missing = Other). Finally at step 3 we entered an interaction term of ethnicity with age. In this instance there was evidence of a significant interaction between age and ethnicity (F(3,140) = 3.2, p = .03). These results suggest that the developmental trends in the magnitude of sex differences on the Progressive Matrices differ across ethnic groups. However, post hoc tests show that it is only the black (N = 6) and south Asian groups (N = 11) that exhibit a significantly different developmental trend from the whites. These samples are small and may reflect nothing more than sample bias. For example, in Africa, dropout rates from school are high, so school samples may be biased in favour of either sex, depending on local customs.

4. Discussion

There are eight points of interest in the results of this meta-analysis. First, it was designed to test the Eysenck-Court-Mackintosh-Jensen hypothesis that there is no sex difference in mean scores on the Progressive Matrices. The results given in Tables 1 and 3 show that this is correct for the Standard Progressive Matrices for the age range 6 to 14 years. However, it is incorrect for the age of 15 years onwards. From this age into adulthood males consistently obtain higher means than females and all of these differences are statistically significant except for the 80-89 age group, for which the difference is of approximately the same magnitude and the lack of statistical significance is attributable to the small sample size. The advantage of boys begins to appear at the age of 14 and increases in size among adults, when it reaches an average for the whole age range of 20-29 through 80-89 of .33d, equivalent to 5 IQ points. It is proposed that this is the best estimate of the male advantage on the Progressive Matrices among adults.

Second, the hypothesis that there is no difference on the Progressive Matrices also fails for the Colored Progressive Matrices for the age range 5 though 11 years shown in Tables 2 and 3. The results show that boys obtain significantly higher means than girls at ages 6, 7, 8 and 11 years, and that for the whole age range boys have an a mean advantage of .21d, equivalent to 3.2 IQ points. However, on the Standard Progressive Matrices over this age range boys have a negligible advantage of .02d.

It is proposed that the explanation for the greater advantage of boys on the Colored Progressive Matrices than on the Standard Progressive Matrices is probably that the Progressive Matrices is not a pure measure of reasoning ability, as has been frequently asserted, but also measures visualization. This was suggested by van der Ven and Ellis (2000) who have shown in a Rasch analysis for unidimensionality of the Standard Progressive Matrices, in a sample of 905 Dutch 12-15 year olds, that the test contains two factors identified as (1) "gestalt continuation" present in Set A and items 1-6 in Set B for which "the correct solution must be found according to some Gestalt continuation rule"; and (2) "analogical reasoning", present in items B 8-12 and in most of the later items. We prefer the word "visualization" to "gestalt continuation" as the widely accepted term for this ability in hierarchical factor models such as that of Carroll (1993). It is known from the meta-analysis of sex differences in spatial abilities carried out by Linn and Peterson (1985) that boys tend to perform a little better on average than girls on visualization. The Colored Progressive Matrices contains more visualization items than the Standard Progressive Matrices, namely most of Set Ab which is absent in the Standard Progressive Matrices. We think it probable that boys perform better than girls on these visualization items and that this is the explanation for the greater male advantage on the Colored Progressive Matrices. This is a hypothesis that needs to be examined.

Three, the results provide tests of two components of the developmental theory of the sex differences in the maturation of intelligence. The component of the theory for the age range 6-14 years is not supported by the results. This states that the sex difference is negligible up to the age of about 8 years, moves in favor of girls at the age of about 9 years as the accelerated growth spurt of girls speeds up their development in a number of characteristics including height, weight and brain size, remains in favor of girls for the next two or three years up to age of around 12 years, following which there is little difference between boys and girls up to the age of 15. The results set out in Table 1 show trends supporting this theory but these are not statistically significant. The second component of the theory stating that from the age of 15 years onwards males obtain higher means than females is confirmed by the results.

Four, the results do not support Mackintosh's (1998) contention that the male advantage has been greater among older cohorts of adults than among young adults. Mackintosh's contention is that among adults the size of the sex difference has declined such that it is negligible in the 20-29 age group even though substantial in older age groups. It can be seen from inspection of the magnitude of the sex differences for the age groups 20-29 through 80-89 set out in Table 2 that there is no tendency for the difference among the younger age groups to be smaller than among older. The sex differences for the 20-49 year olds and for the 60-89 year olds are virtually identical at 0.34 and 0.33, respectively. Mackintosh's (1998) hypothesis is refuted more formally by the regression analysis, which showed no significant effect of generational age on the magnitude of the sex difference in scores on the Progressive Matrices. According to Hunter and Schmidt's 75% rule, there was evidence of moderator variables for children and adolescents of 8 to 18 years of age, though not among adults, except for an anomalous finding for the groups aged 30 to 39 years. However, our test for moderator variables using multiple-regression excluded generational effects as a possible moderator.

Five, the conclusion that from the age of 15 years males have higher mean scores than females has implications for the question of sex differences in general intelligence, fluid intelligence (Gf) and Spearman's g. Both Jensen and Mackintosh draw significant conclusions from their reading of the evidence that there is no sex difference on the Progressive Matrices. Jensen (1998, p.541) argues that, because the Progressive Matrices is the best measure available of Spearman's g, it follows that there is no sex difference in Spearman's g. Mackintosh (1998, p.538) draws a similar conclusion that, because the Progressive Matrices is an excellent measure of Gf (fluid intelligence) and that Gf can be identified with "general intelligence", there is no sex difference in either Gf or in general intelligence. The results of our meta-analysis show that if these identifications are correct, males from the age of 15 years onwards have higher average g, fluid intelligence and "general intelligence" than females by approximately 5 IQ points.

Six, we now consider the issue of possible cross-cultural variation in growth curves. There was some evidence that the patterns in growth curves of sex differences on the Progressive Matrices are culture specific. However, it is evident that all cultures examined in the analysis converge on the same end point by late adolescence and early adulthood. It should be noted that nine of the ten studies of adults from the age of 20 to 80 are from economically developed nations. The remaining study is from Brazil and the sex difference of .28d in the Brazil sample is a little less than .33d in the whole of the ten studies.

Seven, while our results support Lynn's (1994, 1998, 1999) developmental theory of sex differences in intelligence, other explanations of the gradual increase in the mean scores obtained by males relative to those of females from age 15 should be considered. First, there is the possibility of differential experience of men and women. In all societies women are predominantly responsible for the care of children (Whiting & Whiting, 1975), and even professional women have been shown to carry out the majority of domestic work (Yogev, 1982). It is may be possible that this specialisation could affect scores on tests of cognitive ability. Lippa (1998) has shown that there are male and females differences on Prediger's (1992) People-Things dimension such that males are more interested in things and females in people. It has been shown further that genetic variation explained 53% of the variance in within-sex differences in gender diagnosticity, a variable highly correlated with the People-Things dimension (Lippa and Hershenberger, 1999).

A possible interpretation of these findings might be that women's genetic predispositions are such as to maximize their satisfaction with their traditional adult roles and these predispositions are subsequently amplified by gender role socialization (Ridley, 1994; Lippa, 2002, Maccoby, 2000). That the resulting differences in experience can influence scores on intelligence tests is suggested by a study of Ackerman, Bowen, Beier & Kanfer, (2001), which showed that social potency, social closeness and traditionalism-worry are all related to lowered scores on general knowledge, a measure of crystallized intelligence (Carroll, 1993), i.e. those who are socially oriented, among whom women predominate, invest their cognitive resources differently from those who are not socially oriented, and this difference accounts for the lower scores on general knowledge.

Whether this difference would affect non-verbal reasoning ability may be more doubtful. However, studies of experts also suggest how experience might influence fluid intelligence (Ericsson & Kintsch, 1995). They have shown that the major factor that appears to differentiate experts at chess and medical diagnosis from those who are average is the existence of long-term memory structures that augment the capacity of working memory. Since Kyllonen and Christal (1990) have provided the adult evidence that fluid intelligence can be roughly equated to the capacity of working memory, a difference in long-term memory structures due to different adult experiences among men and women may, by augmenting working memory capacity, be responsible for an observed difference in scores on fluid intelligence. The above theory, which posits the joint action of biological and social factors as proximal causes of gender differences, corresponds to what Feingold (1994) describes as a biosocial model.

While differential experience may provide an explanation for the adult sex differences in scores on the Progressive Matrices, there are two aspects of the data that seem incompatible with this. First, if differential experience comprises the underlying mechanism, we would be expect cross-cultural differences sex differences in scores on the Progressive Matrices, since adult sex roles differ across cultures. Yet we did not find this. Second, given the large changes in the adult roles of women in Western industrialized societies post-war, we should expect a reduction in the sex difference over time, yet we found no evidence of generational change.

Eight, whatever the causes of the observed pattern of sex differences on the Progressive Matrices, it would seem desirable to place the overall difference of .33d in adults in context. There is considerable recent evidence that from a position that obtained about 20 years ago, when boys and men consistently outperformed girls and women in terms of educational performance, on many indices girls and women are increasingly either equalling or overtaking boys and men. For example, in the United Kingdom, in 1980, 13 % of young women obtained two or more "A" levels or their equivalent compared with 14% of young men (Ramprakash & Daly, 1983), whereas by the year 2000 the respective figures were 39% and 31% in favour of young women (Matheson & Babb, 2002). In terms of higher education, in 1980 only 37% of first degrees were obtained by women (Ramprakash & Daly, 1983), but by 2001 this figure had risen to 56% (Matheson & Babb, 2002). Up to 1997 men obtained more first class honours degrees that women but in 1998 women for the first time gained more first class honours degrees than men (Higher Education Statistics Agency, 1998). Given that this increasing female advantage in educational achievement coexists with somewhat lower scores among adult women on the Progressive Matrices, it can be inferred that there are other factors predominantly possessed by women that facilitate this achievement. Possibly this may be stronger work motivation. Thus, it has been found in the United States that women obtain lower mean scores on the SAT-M but they did not obtain lower maths grades (Wainer & Steinberg, 1992). The most probable explanation is that women's stronger work motivation compensates for their lower test scores.

Wednesday, March 08, 2006

Book review of Race Differences in Intelligence: An Evolutionary Analysis (2006) by Richard Lynn

Publisher: Atlanta, Georgia: Washington Summit Books (PO Box 3514, Augusta, GA 30914) ISBN 1-59368-020-1 pp. 318., US$37.95 (HB), $20.95 (PB) (add $6 for overseas orders). 

(A review by Prof.J.P.Rushton published in March in Personality and Individual Differences describes the book and is given below)

Lynn’s book represents the culmination of more than a quarter of a century’s work on race differences in intelligence. It was in 1977 that he first ventured into this field – some would say minefield – with the publication of two papers on the IQ in Japan and Singapore. Both showed that the East Asians obtained higher means than white Europeans in the United States and Britain. These initial studies were criticised, but the present book lists 60 studies of the IQs of indigenous East Asians all of which confirm the original claim.

Hitherto studies of race differences in intelligence have been largely conducted and discussed in local contexts. In the United Sates they have been largely concerned with the IQs of whites, blacks, Hispanics, Asians and Native American Indians. In Australia they have been concerned with the low IQ of the Aborigines, and in New Zealand with the low IQ of the Maoris. These differences have typically been explained by racism and discrimination of Europeans against minorities the legacy of slavery, although a number of writers have posited a significant genetic factor (Jensen, 1998; Rushton and Jensen, 2005). Lynn’s book differs in taking a global perspective and consists of a review more than 500 studies published world wide from the beginning of the twentieth century up to the present. He devotes a chapter to each of ten races, differentiated by Cavalli-Sforza, Menozzi and Piazza (1994) into “genetic clusters”, which he regards as a transparent euphemism for races.

His conclusions are that the East Asians (Chinese, Japanese and Koreans) have the highest mean IQ at 105. These are followed by the Europeans (IQ 100). Some way below these are the Inuit (Eskimos) (IQ 91), South East Asians (IQ 87), Native American Indians (IQ 87), Pacific Islanders (IQ 85), South Asians and North Africans (IQ 84). Well below these come the sub-Saharan Africans (IQ 67) followed by the Australian Aborigines (IQ 62). The least intelligent races are the Bushmen of the Kalahari desert together with the Pygmies of the Congo rain forests (IQ 54).

After the ten chapters setting out the evidence for each of the ten races there follows a chapter on the reliability and validity of the measures. These show that the studies have high reliability in the sense that different studies of racial IQs give closely similar results. For instance, East Asians invariably obtain high IQs, not only in their own native homelands but in Singapore, Malaysia, Hawaii and North America. To establish the validity of the racial IQs he shows that they have high correlations with performance in the international studies of achievement in mathematics and science. Racial IQs also have high correlations with national economic development, providing a major contribution to the problem of why the peoples of some nations are rich and others poor. He argues further that the IQ differences between the races explain the differences in achievement in making the Neolithic transition from hunter-gathering to settled agriculture, the building of early civilizations, and the development of mature civilizations during the last two thousand years.


Lynn tackles the problem of the environmental and genetic determinants of race differences in intelligence and concludes that these contribute about equally to the phenotypic differences. He argues that the consistency of racial IQs in many different locations can only be explained by powerful genetic factors. He works out the genetic contribution in most detail for the sub-Saharan Africans. His argument is that sub-Saharan Africans in the United States experience the same environment as whites, as regards determinants of intelligence. He argues that they have as good nutrition as whites, as shown by their having the same average height in studies going back to World War 1, and they have approximately the same education as whites. He presents evidence that blacks in the southern states have very little white ancestry and have an IQ of about 80, and that proposes that this can be adopted as the genotypic IQ of blacks, i.e. the IQ that blacks attain when they are reared in the same environment as whites. The IQ of blacks in sub-Saharan Africa is a good deal lower at 67. Hence, the adverse environment in sub-Saharan Africa, which he regards as consisting principally of poor nutrition and health, contributes about 13 IQ points to the low IQ in sub-Saharan Africa. Lynn’s estimate is not too different from that advanced in 1969 by Jensen to the effect that about two thirds of the low IQ of blacks in the United States is attributable to genetic factors, and the more recent estimate of Rushton and Jensen (2005) that the figure is around 80 percent. Lynn has (unsurprisingly for those familiar with his work) put a bit more weight on the genetic factor.

The last three chapters are concerned with the book’s subtitle (An Evolutionary Analysis) and discusses how race differences in intelligence have evolved. He begins by putting the problem in context by summarizing Jerison’s (1973) classic study showing that during the course of evolution species have evolved greater intelligence in order to survive in more cognitively demanding environments. For instance, in one of the most dramatic of these developments, early mammals evolved larger brains and greater intelligence to survive in the nocturnal environment, for which they needed to evolve larger auditory and olfactory analysing centres in the brain.

The same principle, Lynn argues, explains the evolution of race differences in intelligence in humans. He elaborates the argument he has advanced over the last fifteen years that the race differences in intelligence have evolved as adaptations to colder environments as early humans migrated out of Africa. In North Africa and South Asia, and even more in Europe and Northeast Asia, these early humans encountered the problems of having to survive during cold winters when there were no plant foods and they had to hunt big game to survive. They also had to solve the problems of keeping warm. These required greater intelligence than was needed in tropical and semi-tropical equatorial Africa where plant foods are plentiful throughout the year. He shows that race differences in brain size and intelligence are both closely associated with low winter temperatures in the regions they inhabit. For instance, he gives a figure of 1282 cc for the average brain size of sub-Saharan Africans, as compared with 1367 cc for Europeans and 1416 cc for Orientals. His analysis relating race differences in intelligence to exposure to low winter temperatures has recently been independently corroborated by Templer and Arikawa (2005).

From time to time Lynn notes anomalies in his theory that require explanations. One of these is that the Europeans have made most of the great intellectual advances and discoveries, while the East Asians, despite having a higher IQ have made relatively few (as extensively documented by Murray, 2003). Lynn proposes the explanation for this may be that the East Asians are more conformist that Europeans and this inhibits creative achievement. He also notes one or two anomalies in his cold winter theory of race differences in intelligence. The most striking of these is that the Inuit have been exposed to the coldest winter temperatures and have evolved large brains, the same average size as that of the East Asians. Yet their IQ is only 91, and this is the IQ obtained by those who attend the same schools as Europeans. To explain this anomaly he proposed that two genetic processes must be assumed to explain the evolution of race differences in intelligence. The first of these is that differences in the frequencies of the alleles for high and low intelligence have evolved between races such that the alleles for high intelligence are more common in the races with the higher IQs and less common in the races with the lower IQs. The early humans that migrated out of Africa and spread throughout the world would have carried all the alleles for high and low intelligence with them, but those who colonized Asia and Europe were exposed to the cognitively demanding problems of survival during cold winters. Many of those carrying the alleles for low intelligence would have been unable to survive during the cold winters and the less intelligent individuals and tribes would have died out, leaving as survivors the more intelligent. This process would have reduced and possibly eliminated the alleles for low intelligence, leaving a higher proportion of the alleles for high intelligence. The more severe the winter temperatures, the greater the selection pressure for the elimination of low IQ individuals carrying low IQ alleles. This process explains the broad association between coldest winter temperatures and IQs and brain size.

He now suggests that there must have been a second genetical process that several new alleles for high intelligence must have appeared as mutations in some races but did not appear in others, and once these had appeared they were never transmitted to other races. These new mutant alleles for high intelligence would have been most likely to appear in large populations because a mutation is a chance genetic event and hence would have been more likely to occur in races with large populations than in those with small. The Inuit comprised only very small populations numbering today around 55,000, so they would be unlikely have had mutations for higher intelligence that have to be assumed in the East Asians and Europeans. Once a new mutant allele for higher intelligence had appeared in the East Asians and Europeans it would have conferred a selection advantage and would have spread throughout the group of around 50 to 80 individuals in which people lived during the hunter-gatherer stage of human evolution. It would then have spread fairly rapidly to adjacent groups because hunter-peoples typically have alliances with neighboring groups with which they exchange mating partners, and it is reasonable to assume that this custom was present for many thousands of years during the evolution of the races. These alliances of groups are known as demes, and a new mutant allele for higher intelligence and which conferred a selection advantage would have spread fairly rapidly through a deme. From time to time matings would take place between demes and by this means new mutant alleles for higher intelligence would spread from one deme to another and eventually throughout an entire race.

However, this would take some considerable time, and Lynn proposes that in 25,000 years, consisting of approximately 1,000 generations, an advantageous allele would be transmitted about 800 miles. Hence, an advantageous allele occurring as a mutant in the region of, say, Beijing, 25,000 years ago would not yet have spread outside China and would take another 50,000 years or so to reach the Inuit peoples of far North East Asia and even longer to cross the Bering Straits into Alaska. In addition, there are geographical barriers of high mountains between the East Asians and the Inuit that would have imposed a further impediment for the new alleles for higher intelligence being transmitted from East Asia northwards. He extends this explanation to the low IQs of the Australian Aborigines and Bushmen. These have only been small populations, so the chance of mutations of high IQ alleles in them would have been low.

To the arguments presented by Jensen (1998) for a substantial genetic determination of the difference in intelligence between blacks and whites in the United States, Lynn adds a more general one. He advances the general principle of evolutionary biology that wherever subspecies, strains or races have evolved in different environments they invariably develop differences in all characteristics for which there is genetic variation as a result of mutations occurring in some subspecies and of adaptations to different environments, and asserts that intelligence cannot be an exception. He concludes witheringly that:

“The position of environmentalists that over the course of some 100,000 years peoples separated by geographical barriers in different parts of the world evolved into ten different races with pronounced genetic differences in morphology, blood groups and the incidence of genetic diseases, and yet have identical genotypes for intelligence, is so improbable that those who advance it must either be totally ignorant of the basic principles of evolutionary biology or else have a political agenda to deny the importance of race. Or both “. So much for the assertion of the American Psychological Association’s task force under the chairmanship of Ulrich Neisser set up to produce a consensus statement on what is known about intelligence that concluded that there is no persuasive evidence for genetic race differences (Neisser et al., 1998). With the publication of Lynn’s book it will never again be possible to make this assertion and retain any credibility.

Over the years Lynn has made a number of important contributions to the field of intelligence. He has written the standard works on the dysgenic processes that have been taking place almost worldwide for the last century and on how these could be addressed (Lynn, 1996, 2001). He has shown that the problem of why some nations are rich and others poor is largely explained by the intelligence of the populations (Lynn and Vanhanen, 2002). He has overturned the century long consensus that there is no sex difference in intelligence by showing that men have a higher average IQ than women by approximately 5 IQ points (Lynn and Irwing, 2004). But I would guess that the present book documenting global race differences in intelligence and analysing how these have evolved will come to be seen as his crowning achievement.


References

Cavalli-Sforza, L.L., Menozzi, P. and Piazza, A. (1994) The History and Geography of Human Genes. Princeton, NJ: Princeton University Press.

Jensen, A .R. (1998) The g Factor. Westport,CT: Praeger.

Jerison, H. (1973) Evolution of the brain and intelligence. New York: Academic Press.

Lynn, R. (1996) Dysgenics: Genetic Deterioration in Modern Populations Westport,CT: Praeger.

Lynn, R. (2001) Eugenics: A Reassessment. Westport, CT: Praeger.

Lynn, R. and Irwing, P. (2004) Sex differences on the Progressive Matrices: a meta-analysis. Intelligence, 32, 481-498.

Lynn, R. and Vanhanen, T. (2002) IQ and the Wealth of Nations. Westport, CT: Praeger.

Murray, C. (2003) Human Accomplishment. New York: Harper Collins.

Neisser, U.et al. (1996) Intelligence: knowns and unknowns. American Psychologist, 51, 77-101.

Rushton, J. P. and Jensen, A. R. (2004) Thirty years of research on race differences in cognitive ability. Psychology, Public Policy and Law, 11, 235-294.

Templer, D.I. and Arikawa, H. (2005) Temperature, shin color, per capita income and IQ: an international perspective. Intelligence, (to appear)

Biography

Richard Lynn was born in 1930. He graduated in Psychology and took his Ph.D. at the University of Cambridge. He has been lecturer in Psychology at the University of Exeter, professor of Psychology at the Economic and Social Research Institute, Dublin, and professor and head of the department of Psychology at the University of Ulster; now professor emeritus. His main work has been on intelligence and personality. His books include Personality and National Character (1972),Dimensions of Personality (1980), Educational Achievement in Japan (1988), Dysgenics: Genetic Deterioration in Modern Populations (1996), Eugenics: A Reassessment (2001) and (jointly with Tatu Vanhanen) IQ and the Wealth of Nations (2002).

Awards he has received include the Passingham Prize (Cambridge University Prize for the best Psychology student of the year) and the US Mensa Awards for Excellence 1985 and 1993 for work on intelligence. See his website for more details.