Talk:Race and intelligence/Draft

Race and intelligence research investigates differences in the distributions of cognitive skills among human races. Debates in popular science and academic research over the possible connection of race and intelligence originally began as a comparison of African Americans and Caucasians in the United States, but were later extended to other races and regions of the world. In the US, intelligence quotient (IQ) tests have consistently demonstrated statistical differences: the scores of the African American population are on average lower than that of the White American population; the Asian American population on average scores higher; Amerinds scores on average fall between Caucasian and African American scores. Similar findings have been reported for populations around the world, most notably in Africa, though these are generally considered far less reliable due to the relative paucity of test data and the difficulties inherent in the cross-cultural comparison of intelligence test scores. The distribution of IQ scores has considerable range - individuals in every racial group may have IQ's that lie anywhere on the spectrum of scores. These difference show primarily in aggregate studies.

There are no universally accepted definitions of either race or intelligence in academia, and many factors that could potentially influence the development of intelligence have been advanced to explain the racial IQ gap. There is general agreement that environmental and/or cultural factors affect individual IQ scores, and it is widely assumed that a significant portion of the racial IQ gap is attributable to such factors, though none are conclusively supported by direct empirical evidence. The more controversial view that a significant portion of the racial IQ gap is ultimately of genetic origin has been advanced by academics such as Arthur Jensen, J. Philippe Rushton and Richard Lynn. This claim met with widespread criticism in the popular media, particularly after the publication of "The Bell Curve", and has not to date gained acceptance by the wider academic community.

The racial IQ gap has remained relatively stable since IQ testing began, although IQ scores as a whole have themselves been subject to change over time. The American Psychological Association has concluded that the racial IQ gap is not the result of bias in the content or administration of tests, but that no adequate explanation of it has so far been given.

History
Differences between the average IQ test scores of population groups have been demonstrated but there has been no agreement on whether these are primarily due to environmental/cultural factors or to hereditary genetics.

In the late nineteenth and early twentieth century, group differences in intelligence were assumed to be due to race and, apart from intelligence tests, research relied on measurements such as brain size or reaction times. By the mid-1930s most psychologists had adopted the view that environmental and cultural factors played a dominant role. In 1969 the educational psychologist Arthur Jensen published a long article reviving the older hereditarian point of view, with the suggestion that eugenics was more likely to increase the average intelligence in the US than remedial education for blacks. His work, publicized by the Nobel laureate William Shockley, sparked controversy amongst the academic community and even led to student unrest. A similar debate amongst academics followed the publication in 1994 of The Bell Curve, a book by Richard Herrnstein and Charles Murray reviving the hereditarian viewpoint once more. It provoked not only the publication of several interdisciplinary books on the environmental point of view, some in popular science, but also to a public statement from the American Psychological Association acknowledging a gap between average IQ scores of whites and blacks as well as the absence of any adequate explanation of it, either environmental or genetic. The hereditarian line of research continues to be pursued by a group of psychologists, some of whom are supported by the Pioneer Fund.

Group differences
Intelligence is most commonly measured using IQ tests. These tests are often geared to measure the psychometric variable g (for general intelligence factor). Other tests that measure g (e.g, the Armed Forces Qualifying Test, SAT, GRE, GMAT and LSAT) also serve as measures of cognitive ability. Several conclusions about these types of tests are now largely accepted:
 * IQ scores measure many of the qualities that people mean by intelligent or smart.
 * IQ scores are fairly stable over much of a person's life.
 * IQ tests predict school and job performance to a degree that does not significantly vary by socio-economic or racial-ethnic background.
 * Intelligence is heritable.
 * Family environment and community culture affect IQ, more so in children than in adults.

Test scores
Most of the evidence of intelligence differences between racial groups is based on studies of IQ test scores, almost always using self-reported racial data. Self-reports have been shown to be reliable  indicators of genetic race to the extent that they match up with genetic clusters derived from mathematical clustering techniques, but these techniques do not determine whether these clusters themselves have any relation to intelligence. There are observed differences in average test score achievement between racial groups, which vary depending on the populations studied and the type of tests used. In the United States, self-identified Blacks and Whites have been the subjects of the greatest number of studies. Black-White average IQ differences appear to increase with age, reaching an average of nearly 17 points by age 24, which is slightly more than 1 standard deviation. Arthur Jensen has found, further, that black and white populations regress to these different means, using Galton’s Law of Ancestral Heredity, and argues that this supports a genetic theory, on the grounds that it is difficult to explain this kind of regression based on environmental factors. Richard Nisbett recognizes the existence of this effect, but believes that it could be produced by environmental factors alone, such as parenting practices and subculture pressures. Further research by Jensen and other researchers using structural equation modeling concluded that a model in which genetic and environmental contributions to the IQ gap are in roughly equal proportions best fit the data.

Gaps are seen in other tests of cognitive ability or aptitude, including university admission exams, military aptitude tests and employment tests in corporate settings. The IQ distributions of other racial and ethnic groups in the United States are less well studied. The few Amerindian populations that have been systematically tested, including Arctic Natives, tend to score worse on average than White populations but better on average than Black populations. East Asian populations score higher on average than White populations in the United States as they do elsewhere.

Racial differences in IQ scores are observed around the world. Lynn's meta-analysis lists East Asians (105), Europeans (102), Inuit (91), Southeast Asians and Amerindians (87 each), Pacific Islanders (85), South Asians/North Africans (84), Non-Bushmen sub-Saharan Africans (67), Australian Aborigines (62) and Bushmen (54). International achievement test scores, including TIMSS and PISA, have also been used to estimate average IQ worldwide with similar results where data is available.

Questions about the quality of the data
However, Lynn's findings are highly contested, and some have alleged that Lynn has falsified and manipulated data. . A review by Nicholas Mackintosh of Lynn's book Race Differences in Intelligence, where the above findings are presented, has also criticized Lynn's occasional manipulation of data, some of it originally collected by the reviewer. He has also questioned Lynn's inference elsewhere that Kalahari bushmen, with an average measured IQ of 54, should be regarded as mentally retarded. He concludes that "'Much labour has gone into this book. But I fear it is the sort of book that gives IQ testing a bad name. As a source of references, it will be useful to some. As a source of information, it should be treated with some suspicion. On the other hand, Lynn's preconceptions are so plain, and so pungently expressed, that many readers will be suspicious from the outset.'"

Debate assumptions and methodology
Race and intelligence research involves debate over the links, if any, between race and intelligence. This research is grounded in two controversial assumptions: Both assumptions are disputed.
 * Race is partially genetically based.
 * Intelligence is quantitatively measurable.

There are several conflicting positions. Some scientists argue that the history of eugenics makes this field of research difficult to reconcile with current ethical standards for science. Other scientists insist that, independent of ethical concerns, research into race and intelligence makes little sense because intelligence is poorly measured and because race is a social construction. According to this view, intelligence is ill-defined and multi-dimensional, or has definitions that vary between cultures. This would make contrasting the intelligence of groups of people, especially groups that came from different cultures, dependent mainly on which culture’s definition of intelligence is being used. Moreover, this view asserts that even if intelligence were as simple to measure as height, racial differences in intelligence would still be meaningless since race exists only as a social construct, with no basis in biology.

Unsurprisingly, almost all scientists actively engaged in research in race and intelligence disagree with these two positions. These researchers fall into two groups: hereditarians and environmentalists. Both argue that, although race and intelligence are fuzzy concepts, they can be operationalized enough to draw conclusions about the connections, if any, between the two. In this research, race is almost always measured via self-identification. Subjects are presented with a set of racial options and allowed to place themselves in one (or more) category, or are placed by an interviewer. The set of categories allowed and the words used to describe them varies from study to study and from country to country. Intelligence is generally measured with some form of IQ test.

Hereditarians argue that genetics explain a significant portion (approximately 50%) of the differences in measured intelligence among human races. Leading scholars of this view include Arthur Jensen, Philippe Rushton, Richard Herrnstein, Linda Gottfredson, Charles Murray and Richard Lynn.

Environmentalists argue that the hereditarians are wrong, and that genetic differences are not an important cause of differences in measured intelligence among human races. Leading scholars of this view include Richard Lewontin, Stephen J. Gould, James Flynn, Richard Nisbett and Stephen Ceci.

Other scientists, although accepting the basic assumptions of the debate, believe that there is currently not enough evidence to determine what part, if any, genetics plays in racial differences.

In theory, the dispute could be resolved with a simple experiment. Select 1,000 fertilized eggs at random from the relevant racial groups, say White, Black and East Asian. Place all 1,000 in identical environments, both pre- and post-natal. This would involved creating three separate societies which would be perfectly equal and as similar as possible to contemporary developed countries. Nutrition, family environment, education, popular culture and all other factors which might influence intelligence would need to be identical. For example, the Nobel Prize winners in the White society would all be White, those in the Black society would all be Black and so on.

Then, after 18 years, give the thousand subjects from each racial group an intelligence test. If the group averages are the same, then the hereditarian hypothesis is refuted. Since ethical and practical issues make such an experiment impossible, researchers in race and intelligence need to rely on simpler and less conclusive experiments and data analysis.

There are two primary methods for controlling factors (education, income and so on) which are correlated with IQ test scores and which co-vary with race The first constrains participant selection so that members of all races are equal on the factor in question. For example, if a researcher thinks education is the explanation for the difference, then she could compare the IQs of only similarly-educated members of each group. Showing that the difference is zero for persons matched on education levels would suggest that education is the cause of the difference. Showing that the difference remains for similarly educated people would make it less likely that education differences across race explain differences in average IQ. The second method is similar to the first, but uses statistics (rather than participant selection) to control the factor.

Debate overview
Richard Nisbett, in replying to hereditarian arguments, structures the debate into several major areas.

Heritability


Heritability is a basic concept in population genetics that measures the degree to which variation among individuals is due to inherited factors. Heritability applies to variation within a population, that is, a group sharing the same environment. This is because environment must be kept constant for heritability to be measured. For example, imagine that the height of "ordinary genetically varied corn" is 100% heritable when grown in a uniform environment. Further imagine that two populations of corn are grown: one in a normal nutrient environment and the other in a deficient nutrient environment. Consequently, the average height of the corn grown in the deficient nutrient environment is less than the average height of the corn grown in the normal environment. In such a scenario, the within-group heritability of height is 100% in both populations, but the substantial difference between groups are due entirely to environmental factors. With respect to the Black-White IQ gap, Jensen suggests that effects associated with racism (both overt and institutionalized racism) might be X-factors. Flynn believes that attributing the B-W gap to the effects of racism is incorrect, because the most plausible ways in which discrimination could affect IQ are themselves common environmental factors. These may include psychological effects such as stereotype threat; biological effects such as poor nutrition, health care and living close to toxic environments; and educational effects such as a lack of good schools. Instead, Flynn and his colleague William Dickens have developed more complicated models to explain the black-white gap in terms of environmental factors. One initial motivation of the Dickens-Flynn theory was Flynn's observation that IQ test scores have been rising over time in countries around the world – termed the Flynn effect. Flynn and others believe an explanation for the Flynn effect may elucidate the cause of the B-W gap. Jensen and others disagree.

Environment


Much of the research on this topic has been conducted by Arthur Jensen and James Flynn. Flynn and Jensen consider two general classes of environmental factors: common environmental factors, which vary both within and between groups; and X-factors, which vary between groups but not within groups. Flynn explains in Race, IQ and Jensen (1980) why common environmental factors are inadequate as an explanation for the IQ gap:

After all, if an environmental factor is potent enough to account for the 15-point performance gap between black and white, and if it varies much from person to person within the black population, it would be extremely odd if it accounted for none of the variable performance within the black population! And if it did, it would of course increase the role of environmental factors in explaining IQ variance and thus lower the h2 (within-group heritability) estimate for blacks. [...] If we seize on SES (socio-economic status) as a between-population explanation, who can deny that there are large differences in SES within black America; if we seize on education, who can deny that blacks differ significantly in terms of quality of education?

The alternative to common environmental factors is the hypothesis that the racial IQ gap can be accounted for by X-factors: factors which vary between groups but not within groups. Jensen and Flynn agree that no X-factors have yet been identified that could account for the racial IQ gap. Jensen believes that under these circumstances, the “default hypothesis” should be that the differences in average IQ between races is caused by the same factors that cause within-group variance in IQ, while Flynn believes that the racial IQ gap is caused by X-factors that have yet to be discovered.

Environmental factors including lead exposure, breast feeding , and nutrition can significantly affect cognitive development and functioning. For example, iodine deficiency causes a fall, in average, of 12 IQ points. Such impairments may sometimes be permanent, sometimes be partially or wholly compensated for by later growth. Comprehensive policy recommendations targeting reduction of cognitive impairment in children have been proposed.

Score convergence
The overall average Black-White gap has reduced by one third over the course of the 20th century. For example, the black men inducted into the US armed forces during World War II averaged about 1.5 standard deviations below their white counterparts. This improvement is also reflected in Black-White differences on school achievement tests, which have shrunk from about 1.2 to about 0.8 standard deviations. Flynn claims that the Black-White gap has reduced throughout the 20th century. However, Murray claims that these improvements may have stalled for people born after the early 1970s.

Flynn effect
Although modern IQ tests are unbiased, average test scores over the last century have risen steadily around the world. This rise is known as the "Flynn effect," named for James R. Flynn, who did much to document it and promote awareness of its implications. The effect increase has been continuous and approximately linear from the earliest years of testing to the present.

This means, given the same test, the mean performance of Blacks today could be higher than the mean for Whites in 1920, though the gains causing this appear to have occurred predominantly in the lower half of the IQ distribution. If an unknown environmental factor can cause changes in IQ over time, then contemporary differences between groups could also be due to an unknown environmental factor.

Nichols (1987) critically summarized the argument as follows:
 * 1) We do not know what causes the test score changes over time.
 * 2) We do not know what causes racial differences in intelligence.
 * 3) Since both causes are unknown, they must, therefore, be the same.
 * 4) Since the unknown cause of changes over time cannot be shown to be genetic, it must be environmental.
 * 5) Therefore, racial differences in intelligence are environmental in origin.

Dickens (2005) states that "Although the direct evidence on the role of environment is not definitive, it mostly suggests that genetic differences are not necessary to explain racial differences. Advocates of the hereditarian position have therefore turned to indirect evidence ... The indirect evidence on the role of genes in explaining the Black-White gap does not tell us how much of the gap genes explain and may be of no value at all in deciding whether genes do play a role. Because the direct evidence on ancestry, adoption, and cross-fostering is most consistent with little or no role for genes, it is unlikely that the Black-White gap has a large genetic component."

African IQ
The very low IQ scores reported for sub-Saharan African populations (average of 70) are controversial.

Spearman's Hypothesis


Spearman's hypothesis asserts that group differences on intelligence test scores are caused primarily by group differences on the general intelligence factor (abbreviated g). The general factor is a statistical construct that measures what is common to the scores of all IQ test items. How well a person does on one IQ sub-test is usually correlated with how well he or she does on other sub-tests. This is the essence of g.

Jensen developed a statistical technique known as the method of correlated vectors to test Spearman's hypothesis. The idea is that a rank ordering of IQ sub-test items by g-loadings should correlate with the magnitude of the race difference on those items, if indeed g is their cause. For example, digit span backward is more g-loaded than is digit-span forward. And, the race difference on the former is about twice as large as the race difference on the latter.

Spearman's hypothesis is not without its critics. Psychologists Hunt and Carlson write write:

One of the most widely cited pieces of evidence (although not the only one) for biological differences in intelligence, sometimes referred to as Spearman's hypothesis (Jensen, 1998), rests on an indirect argument constructed from three facts. The first is that various IQ measures are substantially correlated, providing evidence for general intelligence. Although tests do vary in the extent of their g loading, factor structures are similar over several test batteries (Johnson, Bouchard, Krueger, McGue, & Gottesman, 2004). The second is that, within Whites, the g factor appears to have a substantial genetic component (see citations in Rushton & Jensen, 2005a). The third fact is that the g loadings of tests are substantially and positively correlated with the difference between the mean White and African American score on each subtest within a battery of tests. This analysis has been referred to as the "method of correlated vectors" (Jensen, 1998). Because it has also been well established that general intelligence has a substantial genetic component, results from the method of correlated vectors have been offered as putative evidence that the "default hypothesis" ought to be that about 50% of the variance in the African American versus White difference reflects genetic differences in a potential for intelligence (Jensen, 1998; Rushton and Jensen, 2005a).

They further summarize criticisms of this position:

Technical objections have been made to the method of correlated vectors and to a somewhat stronger condition: that if the within-group correlations between measures are identical across groups, the between-group differences must arise from the same cause as the within-group correlations (Widaman, 2005). The essence of these objections is that the method of correlated vectors does not consider alternative hypotheses concerning the latent traits that might give rise to the observed difference in test scores. When a more appropriate method of analysis, multigroup confirmatory factor analysis, is applied, it has been found that Spearman's hypothesis (i.e., that the difference is due to differences in general intelligence) is only one of several models that could give rise to the observed distributions in test scores (Dolan, 2000). These findings render the method of correlated vectors ambiguous—which is not the same as saying that the Jensen-Rushton position is incorrect. Our point is that the argument for the default hypothesis is an indirect one. It would be far better if a direct causal argument could be made linking racial/ethnic genetic differences to studies of the development of the brain.

Brain size
On average, the brains of African-Americans are 5% smaller than the brains of Whites and 6% smaller than East Asians, according to studies of brain weight at autopsy, endocranial volume of empty skulls, head size measurements by the U.S. military and NASA, and two dozen MRI volumetric studies. Proponents of both the environmental and hereditarian perspective believe that this variation is relevant to the racial IQ gap, although they disagree as to its cause. Ulric Neisser, The Chair of the APA’s Task Force on intelligence, acknowledges the brain size difference, but points out that brain size is known to be influenced by environmental factors such as nutrition, and that this fact has been demonstrated experimentally in rats. He thus believes that data on brain size cannot be considered strong evidence for a genetic component to the IQ difference. Rushton and Jensen disagree, citing several studies of malnourished East Asians showing that they have larger brains than Whites, and studies demonstrating the brain size difference at birth and prenatally just a few weeks after conception.

A third perspective is offered by Leonard Lieberman, who believes that human variation in brain size is primarily genetic and an adaptation to climate, but that this variation should be viewed as being based on biogeographic ancestry and independently of “race”.

Much of the research into the neuroscience of intelligence has involved indirect approaches, such as searching for correlations between psychometric test scores and variables associated with the anatomy and physiology of the brain. Historically, research was conducted on non-human animals or on postmortem brains. More recent studies have involved non-invasive techniques such as MRI scans as they can be conducted on living subjects. MRI scans can be used to measure the size of various structures within the brain, or they can be used to detect areas of the brain that are active when subjects perform certain mental tasks.

In a study of the head growth of 633 term-born children from the Avon Longitudinal Study of Parents and Children cohort, it was shown that prenatal growth and growth during infancy were associated with subsequent IQ. The study’s conclusion was that the brain volume a child achieves by the age of 1 year helps determine later intelligence. Within human populations, studies conducted to determine whether there is a relationship between brain size and a number of cognitive measures have "yielded inconsistent findings with correlations from 0 to 0.6, with most correlations 0.3 or 0.4."

A study on twins showed that frontal gray matter volume was correlated with g and highly heritable. A related study has reported that the correlation between brain size (reported to have a heritability of 0.85) and g is 0.4, and that correlation is mediated entirely by genetic factors.

Reaction time
Reaction time (RT) is the elapsed time between the presentation of a sensory stimulus and the subsequent behavioral response by the participant. RT is often used in experimental psychology to measure the duration of mental operations, an area of research known as mental chronometry. In psychometric psychology, RT is considered to be an index of speed of processing. That is, RT indicates how fast the thinker can execute the mental operations needed by the task at hand. In turn, speed of processing is considered an index of processing efficiency. The behavioral response is typically a button press but can also be an eye movement, a vocal response, or some other observable behavior.

Scores on many but not all RT tasks tend to correlate with scores on paper and pencil IQ tests. This is especially true for so-called elementary cognitive tasks (ECTs). These require participants to perform trivially simple cognitive tasks, like deciding which of two briefly-presented lines is longer (the inspection time task), or which of three lighted buttons is farthest away from the other two (the odd man out task).

Most people can perform ECTs with near 100% accuracy, but individual differences in RT on these tasks are large and correlate well with IQ scores. Jensen (2001) argues that ECTs could replace traditional IQ tests as measures of intelligence, because the former are measured on a ratio scale whereas IQ tests only rank people on an ordinal scale. Jensen has invented a Jensen box to present ECT task stimuli to participants in a precise, standardized fashion. Not all RT tasks, however, are good measures of intelligence. In general, RT on tasks that take between 200 milliseconds and 2 seconds to perform tend to correlate well with IQ. Tasks that most people can do faster than 200 milliseconds generally measure the efficiency of sensory processes (seeing, hearing) rather than intelligence. Tasks that take longer than about 2 seconds typically allow for strategic differences among people which cloud any relationship between RT and IQ (for these tasks, accuracy-- versus speed-- is likely more related to IQ).

Reaction time best predicts IQ test scores when participants perform many trials (i.e., 100s) of the same ECT. Aggregating average reaction times across different ECTs also produces significantly larger RT/IQ correlations. In many studies, the within person variability of RT is also a strong predictor of IQ. Participants showing relatively large RT differences from trial to trial tend to score lower on IQ tests than do participants who do not deviate much in their reaction time from trial to trial. Finally, the slowest trials for any person tend to better predict that person's IQ relative to either his or her average or fastest response.

Although the literature on RT is vast, far less research has looked at race differences on RT as a potential explanation for the race/IQ gap. The general pattern, however, is that race differences exist on ECT performance, and that these differences are in line with those found on traditional IQ tests. For example, a recent study in the journal Intelligence looked at race differences on the Wonderlic Personnel Test (a traditional paper and pencil IQ test) and performance on two ECTs (an inspection time and choice reaction time task). A black/white difference was found on the Wonderlic, and both ECTs. Statistical mediation was found in that controlling for race differences on the ECTs resulted in the race difference on the Wonderlic no longer being significant.

Adoption studies
The Minnesota Transracial Adoption Study examined the IQ test scores of 130 black/interracial children adopted by advantaged White families. The aim of the study was to determine the contribution of genetic factors to the poor performance of black children on IQ tests as compared to White children. The following table provides a summary of the results.

Policy relevance
In response to criticism that their conclusions would have a negative effect on society if they were to gain wide acceptance, Jensen and Rushton have justified their research in this area as being necessary to answer the question of how much racism should be held responsible for ethnic groups' unequal performance in certain areas. They maintain that when racism is blamed for disparities which are the result of biological differences, the result is mutual resentment, and unjustified punishment of the more successful group. They state: [T]he view that one segment of the population is largely to blame for the problems of another segment can be even more harmful to racial harmony, by first producing demands for compensation and thereby inviting a backlash. Equating group disparities in success with racism on the part of the more successful group guarantees mutual resentment. As overt discrimination fades, still large racial disparities in success lead Blacks to conclude that racism is not only pervasive but also insidious because it is so unobservable and "unconscious." Whites resent that nonfalsifiable accusation and the demands to compensate blacks for harm they do not believe they caused.