IQ and Testing: Critiques
IQ and Testing: Critiques
IQ and Testing: Critiques
A variety of critiques have been offered by eminent scholars regarding the use of IQ tests to draw conclusions about racial/ethnic (particularly black-white) group differences. Before detailing these critiques, however, it is necessary to consider the nature of the argument about IQ and race to which these rebuttals respond. Briefly, the main argument is that racial differences in IQ equate to innate racial differences in intelligence. Proponents of this position include Arthur Jensen, J. Philippe Rushton, Richard Herrnstein, Charles Murray, and Richard Lynn. Their position can be summarized as follows: (1) The black-white IQ gap is generally about 15 points on a standard IQ test (one standard deviation), (2) the IQ tests used are equally fair and valid measures of actual intellectual ability in both blacks and whites, (3) differences in IQ are largely genetic in origin, and (4) the 15-point gap cannot be explained by environmental factors, such as whites’ greater access to high-quality schooling, nutrition, health care, and overall economic advantage.
This notion of a genetic or inherent inferiority of blacks is then extended to explain the less optimal living conditions of blacks, within individual countries as well as across countries, with black communities and entire African nations being seen as economically lacking due to the
inherent intellectual deficiencies of their citizens. Another way to think of this argument is within the “nature-nurture” paradigm in science. The proponents of innate differences argue that the observed differences are due to genes or nature, whereas the opponents of this view argue that observed differences are a result of environmental deprivation, poverty, and racism. What follows is a selection of rebuttals to the “innate differences in intelligence” viewpoint, (this is not an exhaustive list, nor does it address the fundamental question of whether IQ tests measure true “intelligence,” whatever that may be).
Anyone who has taken an IQ test (or related tests, such as the SAT or the GRE) recognizes that the types of questions on the test may be more familiar to some people than to others. Questions from the Wechsler Intelligence Scale for Children, 3rd ed. (WISC-III) referring to, for example, “advantages of getting news from a newspaper rather than from a television news program” (Wechsler, p. 138), “why it is important for cars to have license plates” (Wechsler, p. 137), “why you should turn off lights when no one is using them” (Wechsler, p. 134), “what is an umbrella?” (Wechsler, p 108), and “in what way are a telephone and a radio alike?” (Wechsler, p. 78), would not be equally difficult, even when translated, for individuals from more and less developed countries, or even for people coming from upper-middle-class, working-class, or impoverished families. To be culturally fair, people must be tested using questions that tap knowledge to which the people have been equally exposed—and which is equally valued in the cultures of these people. Unless test equivalence is assured, comparisons across people and groups of people with differing backgrounds can be meaningless.
In response to the above argument, it has been suggested that assessments of IQ that are not as overtly culturally and linguistically bound as the WISC-III, should be used. An example of a potentially more culturally fair test is called the Raven Progressive Matrices, which relies on complex geometric shapes and pictures to assess IQ. However, the cultural neutrality of these tests may be illusory. In a review of potential environmental causes for the worldwide systematic increase in IQs for all developed nations, Wendy Williams noted in 1998 that the contemporary visual world offers many children mazes and games on the backs of cereal boxes and on placemats at fast food restaurants, in addition to their omnipresence on the computer. For children not exposed to such stimuli, the Raven tests may be a much less familiar—and thus more difficult—experience. In fact, a 1998 review of the literature by Nicholas Mackintosh noted that there is “no reason to suppose that the ability to solve arbitrary abstract problems, such as those found in Raven’s tests, is any less a learnt skill than the ability to do mental arithmetic or answer questions about the meanings of words” (Mackintosh 1998, p. 171). Mackintosh cites a study (Sharma 1971) that showed that children’s scores on Raven’s Matrices varied as a function of how long they had been resident in Britain: Those still resident in India and those from the same district in India living in Britain for less than two years scored in the low 80s, while those originally from the same district who had been resident in Britain for more than six years scored more than 100 (Mackintosh 1998, pp. 171–172). Thus, the Ravens, like language-based IQ tests, may suffer from cultural bias.
An interesting parallel to the African-American situation is the historical experience of Caucasian immigrants to the United States. Richard Lynn’s data from 1978 show that every ethnic group, when tested upon entrance to the United States, scored relatively poorly— approximately one standard deviation below the mean. This is true for immigrants from India, Yugoslavia, Greece, Spain, eastern Europe, southern Europe, Portugal, Iran, and Iraq. Subsequent generations of their offspring, however, show an increase in test scores, to the point where they equal or exceed the national average of the host country. In short, these immigrants came in with mean IQs of 85, while their children and their children’s children have mean IQs of 103 (for people from India, for example). Clearly, this increase in scores is due to the effects of environmental factors (including cultural change and test familiarity) and not genetic factors.
On a broader note, the very concept of intelligence varies from place to place. For people from a culture that values scoring high on IQ tests, taking such a test is a different matter than it is for people who do not value high test scores. Robert Sternberg and colleagues summed up a discussion of the cross-cultural validity of IQ tests: “Scores from tests used in cultures and sub-cultures other than those for which the tests were specifically created are suspect, and probably of doubtful validity in many if not most cases” (Sternberg, Grigorenko, and Bundy 2001, p. 29).
Studies used to support claims of genetic causation of IQ differences often confound nature and nurture. This is a problem in adoption studies, both cross-race and within-race. Some studies of black and Asian infants reared by white families show intelligence consistent with the child’s race rather than the race of the adoptive family, arguably providing support for innate racial differences. But there are problems with designing a perfect experiment using real adopted children. First, it is difficult to ensure that adoptive homes are truly randomly assigned and that the children are representative. For example, are adoption rights enforced at the same time (e.g., at birth or at identical ages) for adoptee-adoptee comparisons? Are the impacts of societal racism on the respective child-ren’s upbringing avoided? In such studies, race and environment are often confounded, rendering conclusions unclear.
Similar problems apply to within-race studies of identical twins adopted apart. The observed high correlation of identical twins has been used to argue for genetic causation, but this viewpoint ignores environmental interaction. Much attention has been paid to the extraordinary similarities of identical twins, even those who were separated at birth and meet up as adults, only to learn that they both collect balls of string and jiggle the toilet handle three times (Bouchard, et al. 1990). While these stories are fascinating and clearly reveal that genes are very important, it is essential to remember the dramatic similarity of the environments identical twins are generally raised in, even when they are adopted by separate families. Economic characteristics of these families are often very similar. They live in comparable communities, for example, and share a common culture—even more so than one might expect, because many adoptions are arranged by religious organizations and social workers who seek similar values and attributes across adoptive families. Heredity surely controls part of how intelligent any one person will be, but extrapolation from these studies to explain racial differences in intelligence is problematic.
The “innate differences in intelligence” argument depends on notions of heritability (percentage of variance explained by genes) of IQ and genetic causality. There is no such thing as a universal “heritability of intelligence,” only the heritability of intelligence in such and such a population at a particular time. For example, the environmental component of variance is likely to be much greater in a sample where some children attend school and some do not, than it is within a sample where schooling is universal (such as the United States). Heritability must therefore be regarded as “sample specific,” varying with population and cohort. Heritability is lower among poor people than wealthy people, for example. Imagine two children, both with innate or genetic gifts for music, growing up in homes with very different economic circumstances. One child is given music lessons and access to musical instruments virtually from the time she can walk; the other child has none of these advantages. The former will thus have the opportunity to develop and display more of her genetically rooted talents, yielding higher heritability.
More generally, the variance of a trait within a group does not predict the variance between that group and another, because the differences in genes and environments within a group do not say anything about the differences in genes and environments between groups. Each measure of heritability applies only to the population from which it came, at a particular time and place.
Contrary to the “innate differences” argument, scores on IQ tests have been demonstrated to be affected by environmental factors, such as education, and by environmental changes over time. For example, exposure to schooling increases IQ, so that the more schooling an individual receives, and the higher the quality of this schooling, the higher the person’s IQ score will be, on average (Ceci 1991). Schooling has been shown to increase IQ in studies of children tested before versus after school vacations (Jencks et al. 1972), of children leaving school early (Harnquist 1968) and starting school late (Schmidt 1967), of children with birthdays separated by a single day but whose number of years in school differs by a whole year due to school admissions cutoffs (Cahan and Cohen 1989), and so on. Schooling and IQ have also been shown to increase individual income, which further contributes to the cycle of wealth resulting in higher IQs (Ceci and Williams 1997). How does schooling exert these positive effects on IQ? Basic familiarity with the types of questions on IQ tests is one key mechanism. Another is that schools specifically train students in the types of abilities that help a person answer IQ test questions correctly. Wealthy countries and wealthy communities obviously have more money to spend on schooling, and these communities tend more often to be white.
IQ test scores also change over time. IQs have been increasing at a steady rate for the last century (Flynn 1987; 1999; 2000). Because test manufacturers change the norms over time and keep resetting the average score to 100, it took some time before anyone noticed that the number of questions the average person was answering correctly was steadily increasing. The Raven Matrices, show the most dramatic increase across generations. For example, there was a gain of 20 IQ points in thirty years for Dutch men. The significance of this worldwide rise in IQ scores (known as the Flynn Effect) to the “innate differences in intelligence” argument is that it provides clear evidence for the strong impact of environment on IQ test scores. Given the economic conditions that African Americans have experienced, dramatic differences in IQ could thus be possible with economic enrichment. As one example, in 1998 Min-Hsiung Huang and Robert Hauser analyzed scores on a vocabulary subtest of the general social survey, in which the exact same vocabulary words have been used over and over, and found that black adults showed the largest gains in scores over time. This finding argues for the dramatic effects of improving environments, access to better schooling, and other changes that have accompanied blacks’ increased economic success. In sum, the Flynn Effect shows that an IQ score is not genetic destiny. Even within a given genotype, there is considerable room for IQs to increase substantially.
As a blue-ribbon panel of experts on intelligence concluded, “Heritability does not imply immutability” (Neisser et al. 1996, p. 86). Even highly heritable traits, such as height, can nevertheless change dramatically due to the environment. For example, the children of Japanese immigrants to the United States have usually been taller than their parents, due to better nutrition. Environment is an omnipresent contributing factor in every situation. For instance, it has been argued that racial IQ differences are due to differences between races in average brain size, and there is indeed a positive correlation between brain size and intelligence. But again, it is not clear whether any such differences are genetic or environmental in origin. It is simply not known if purported racial differences in brain size or IQ would be eradicated by equalization of environments.
Finally, the notion that there are “pure gene pools” for blacks versus whites ignores the biological reality that humans are a blended species, and that they are becoming even more so. Attempts to explain IQ differences using blood markers for African versus European ancestry have been unconvincing, revealing only a negligible relationship between IQ and European genes, which is itself potentially attributable to differential treatment of lighter-skinned versus darker-skinned blacks—an environmental effect itself (Nisbett 2005). Some proponents of the “genetic differences in intelligence” argument rely on evidence linking observed “racial characteristics” with underlying genetic differences, suggesting that, if races differ on other aspects such as skin and hair color, it is likely that they would differ on the genes for intelligence. However, genetic differences on one dimension imply nothing about differences on others, particularly when some traits under consideration are caused by individual genes and others are polygenic (i.e., caused by multiple genes—a classic example being intelligence, a broadly polygenic trait). In sum, the picture of how IQ test scores are used to make comparisons between races is often unfocused or muddled, and it suffers from inaccurate and incomplete reasoning on multiple dimensions.
SEE ALSO Heritability.
Bouchard, Thomas J., Jr., et al. 1990. “Sources of Human Psychological Differences: The Minnesota Study of Twins Reared Apart.” Science 250 (4978): 223–228.
Cahan, Sorel, and Nora Cohen. 1989. “Age versus Schooling Effects on Intelligence Development.” Child-Development 60 (5): 1239–1249.
Ceci, Stephen J. 1991. “How Much Does Schooling Influence General Intelligence and Its Cognitive Components? A Reassessment of the Evidence.” Developmental Psychology 27 (5): 703–722.
———, and Wendy M. Williams. 1997. “Schooling, Intelligence, and Income.” American Psychologist 52 (10): 1051–1058.
Flynn, James R. 1987. “Massive IQ Gains in 14 Nations: What IQ Tests Really Measure.” Psychological Bulletin 101: 171–191.
———. 1999. “Searching for Justice: The Discovery of IQ Gains Over Time.” American Psychologist 54: 5–20.
———. 2000. “IQ Gains, WISC Subtests and Fluid G: G Theory and the Relevance of Spearman’s Hypothesis to Race.” In The Nature of Intelligence (Novartis Foundation Symposium 233), edited by Gregory R. Bock, Jamie A. Goode, and Kate Webb, 202–227. Chichester, U.K.: Wiley.
Grieco, Elizabeth M., and Rachel C. Cassidy. 2001. “Overview of Race and Hispanic Origin.” In Census 2000 Briefs. Washington, DC: U.S. Census Bureau.
Harnquist, Kjell. 1968. “Relative Changes in Intelligence from 13 to 18.” Scandinavian Journal of Psychology 9 (1): 50–64.
Herrnstein, Richard J., and Charles Murray. 1994. The Bell Curve: Intelligence and Class Structure in American Life. New York: Free Press.
Huang, Min-Hsiung, and Robert M. Hauser. 1998. “Trends in Black-White Test-Score Differentials: The WORDSUM Vocabulary Test.” In The Rising Curve: Long Term Gains in IQ and Related Measures, edited by Ulric Neisser, 303–332. Washington, DC: American Psychological Association.
Jencks, Christopher, et al. 1972. Inequality: A Reassessment of the Effects of Family and Schooling in America. New York: Basic Books.
Lynn, Richard 1978. “Ethnic and Racial Differences in Intelligence: International Comparisons.” In Human Variation: The Biopsychology of Age, Race, and Sex, edited by R. Travis Osborne, Clyde E. Noble, and Nathaniel Weyl. New York: Academic Press.
———, and Tatu Vanhanen. 2002. IQ and the Wealth of Nations. Westport, CT: Praeger.
Mackintosh, Nicholas J. 1998. IQ and Human Intelligence. Oxford: Oxford University Press.
Neisser, Ulric, et al. 1996. “Intelligence: Knowns and Unknowns.” American Psychologist 51 (2): 77–101. Available from http://www.michna.com/intelligence.htm.
Nisbett, Richard E. 2005. “Heredity, Environment, and Race Differences in IQ: A Commentary on Rushton and Jensen.” 2005. Psychology, Public Policy, and Law 11 (2): 302–310.
Plomin, Robert, John C. DeFries, and Gerald E. McClearn. 1980. Behavioral Genetics: A Primer. San Francisco, CA: W. H. Freeman.
Renner, Michael J., and Mark R. Rosenzweig. 1987. Enriched and Impoverished Environments: Effects on Brain and Behavior. New York: Springer-Verlag.
Rushton, J. Philippe, and Arthur R. Jensen. 2005. “Thirty Years of Research on Race Differences in Cognitive Ability.” Psychology, Public Policy, and Law 11: 235–294.
Sharma, R. 1971. “The Measured Intelligence of Children from the Indian Subcontinent,” Ph.D. diss., University of London (cited in Mackintosh 1998).
Schmidt, W. H. O. 1967. “Socio-Economic Status, Schooling, Intelligence, and Scholastic Progress in a Community in Which Education Is Not Yet Compulsory.” Paedogogica Europa 2: 275–286.
Sternberg, Robert J., Elena L. Grigorenko, and Donald A. Bundy. 2001. “The Predictive Value of IQ.” Merrill-Palmer Quarterly 47 (1): 1–41.
Williams, Wendy M. 1998. “Are We Raising Smarter Children Today? School- and Home-Related Influences on IQ.” In The Rising Curve: Long-term Changes in IQ and Related Measures, edited by Ulric. Neisser. Washington, DC: American Psychological Association Books.
Wechsler, David. 1991. Manual for the Wechsler Intelligence Scale for Children, 3rd ed. San Antonio, TX: The Psychological Corporation.
Wendy M. Williams
Susan M. Barnett
Jeffrey M. Valla