Cernovsky, Z. Z. (1997)
A critical look at intelligence research,
In Fox, D. & Prilleltensky, I. (Eds.) Critical Psychology, London: Sage, ps 121-133.
Psychology's support for an unjust status quo has taken many forms over the years. As noted in several chapters in this book, one important form is the use of psychological tests to explain - in effect, to justify - a lack of societal equality. Administering and interpreting intelligence tests are primary examples of how the routine work of mainstream psychologists can hinder progress toward social justice. Because psychological tests are research tools," the public often assumes that their results or interpretations are infallible. This assumption is a mistake, one with serious negative consequences for members of groups seen as genetically inferior" as "proven" by "science."
Psychologists, educators, psychiatrists, and the lay public often misunderstand intelligence as consisting exclusively or primarily of the skills assessed by modem tests of intelligence. These IQ (intelligence quotient) tests have an excessively narrow focus on skills and tasks acquired and rehearsed in the process of formal or informal schooling. The narrow focus prevents these tests from detecting other crucial ingredients of intelligence such as creativity or social intelligence (the ability to understand, accurately perceive, and influence emotional states and social behavior of others).
These IQ test limitations have historical roots. The first generally accepted intelligence test was devised for school settings by Binet and Simon in France at the dawn of the twentieth century. This test assessed skills expected of school children of different ages. Its purpose was to serve as an objective criterion for sifting out children considered retarded (to be placed in special classes) and for placing other children at their appropriate grade level (see historical background in Fancher, 1985). Binet's test items were later extended for use with adults.
Today, the items of adult IQ tests still remain pervasively based on school-related skills. The test taker and the test administrator usually follow roles analogous to those of a school teacher and a pupil at exams. These roles encompass various implicit requirements specific to Western culture, as noted below.
In this chapter, I discuss underlying conceptual and methodological issues in contemporary research on human intelligence. A major obstacle in intelligence research is the use of inadequate measurement tools. Another obstacle is the frequently encountered static view of intelligence as biologically transmitted via genes and relatively immutable (unchangeable) during the lifetime. Many psychologists also treat intelligence as a unidimensional phenomenon, reducible to a single IQ measure. This unidimensional approach fails to adequately explore creativity, one of the most vital aspects of intelligence. Contemporary fashionable "academic" trends such as J. Philippe Rushton's and Richard Lynn's reports on genetically based "racial differences" in intelligence are often based on an archaic or incompetent methodology. Those who follow these trends will miss the opportunity to more fully use the human potential for economic and emotional benefit of all involved.
The intelligence test interpreter typically assumes that the test taker is motivated to make the best possible effort. However, this motivation to succeed and excel may be relatively weak or even absent in some segments of our society. In addition, test items traditionally are imbued rather exclusively with the mainstream culture of the country where the test was developed. In some cases, intelligence tests measure the extent to which the individual is familiar with that particular cultural tradition rather than the test taker's cognitive talents per se. For instance, questions such as "Who was the American president during the Second World War?" have limited validity for assessing intelligence outside North American mainstream culture. Items assessing arithmetic skills or rote memory (memorising series of digits comparable to telephone or fax numbers or postal codes) may also be less relevant in cultures where numeric coding is less pervasive.
Some psychologists misinterpret tests such as the Raven's Progressive Matrices as measures of fluid intelligence, relatively independent of formal schooling. Yet the Raven's Matrices obviously contain items that can be solved more easily by persons trained in geometry or in algebraic formulae involving negative and positive numbers. These tests are of dubious value if administered to persons raised within a dramatically dissimilar culture. The traditional Eskimo, the rural Chinese, Tibetans from remote and relatively inaccessible villages, disadvantaged South African blacks, or US blacks and Latinos isolated in a ghetto would likely perform better on items whose content is more consistent with their well rehearsed normal daily activities.
Multiple subcultural discrepancies between the test taker and the test developer or test interpreter often prevent IQ scores from adequately reflecting human intellectual potential. Some of these discrepancies were extensively discussed by Dalton Miller-Jones (1989). Noteworthy is the particular use of language in specific sociocultural environments. For example, in response to the item "What is a hat?," the child raised in a poverty stricken ghetto subculture may reply with action-related elementary words such as "you put it on your head." This response is more adaptive in the particular subculture than a dictionary or encyclopedia style definition expected by IQ tests. The dictionary style response could lead to misunderstandings or provoke the disapproval of significant others.
Classical textbooks of psychological testing have warned that "it is unlikely that any test can equally be fair to more than one cultural group” (Anastasi, 1988: 357). Numerous academic psychologists still misinterpret IQ scores from other cultures as indicating genetic inferiority of these groups. For example, Richard Lynn extensively relies, in his postulate of racial inferiority of blacks, on tests such as the Raven's Matrices used in African settings (see critique by Leo Kamin, 1995). Those who defend Lynn's position may argue that blacks are in poverty stricken ghettos because of their genetically inferior intelligence and that their lifestyle and economic condition are a consequence of inborn intellectual inferiority.
This circular reasoning ignores the value of high-quality formal schooling or the impact of conveniences available rather exclusively for the upper socioeconomic class, such as CD-ROM equipped computers or frequent travel to other countries.
These misinterpretations of scores on IQ tests have a long history. Some of the psychologists involved in assessing immigrants on their arrival to the United States in the first half of this century interpreted low test scores as a sign of feeble-mindedness The prominent psychologist Henry Goddard reported extremely high rates of mental retardation for several immigrant groups (Jews, Italians, and Russians) and concluded that "One can hardly escape the conviction that the intelligence of the average third class immigrant is low, perhaps of the moron grade" (1917: 243). Similar misleading claims typically have been associated with a reliance on poor methods and blatantly nonrepresentative samples. For example, Goddard himself admitted that his data were only from six small highly selected groups. Many of these immigrants were not sufficiently familiar with English or psychological tests to comprehend what was happening during the assessments. Psychologists naively considered non-verbal instructions for these IQ tests as adequate, on the untested assumption that non-verbal communication does not differ from culture to culture.
IQ Scores as Indicators of Criminality?
As mentioned by Evans and Waites (1981), American psychologists often misinterpreted IQ scores as an indicator of criminal potential. According to Terman, "all feeble-minded are at least potential criminals. That every feeble-minded woman is a potential prostitute would be hardly disputed by any one. Moral judgement, like business judgement, social judgement, or any other kind of higher thought process, is a function of intelligence" (1919: 11). Goddard's (1917) article on feeble-mindedness among non-Anglo-Saxon immigrants to the United States may have contributed to the subsequent application of very restrictive immigration quotas. Countless refugees and dissidents in German Nazi or Soviet dominated territory perished during World War II and the postwar decades as a result of these quotas.
Heritability of IQ Scores?
Both genetic factors and environmental factors may determine the level, style, and content of adult intellectual functioning. The environmental influences include both psychosocial influences (such as parental stimulation, formal education, peer pressure) and biological influences (such as nutrition, climate, bacteria, and environmental toxins). There have been numerous attempts to determine the relative contribution of the genetic and environmental factors. Despite these efforts, no scientist has succeeded so far in offering a generally accepted theoretical solution for determining the extent of heritability of intelligence. Published estimates of its heritability vary between the statistically feasible extremes. However, the underlying procedures are based on false premises and dubious statistical methodology (see Crusio, 1990; Roubertoux and Capron, 1990a; 1990b; Schonemann, 1989; 1990; 1992; 1995; Taylor, 1980). The case of the British psychologist Sir Cyril Burt is instructive.
During and after World War II, Burt was revered as an authority on the question of heritability of intelligence. Favouring hereditary over environmental explanations, he supported his opinion with data from studies comparing IQ score similarity of monozygotic (i.e., identical) twins with their genetically less closely related siblings. Burt's particular merit seemingly consisted in gathering the largest set of data on identical twins who were raised separately. According to Burt, these twins were separated in early childhood and raised in different socioeconomic environments. His data, if credible, would have helped to tentatively explore the relative contribution of environments and of genes to adult human intelligence (for historical background, see Fancher, 1985; Hearnshaw, 1979).
Burt's methodology was generously and uncritically praised by several politically influential figures within academic psychology such as Hans Eysenck and Arthur Jensen. After Burt's death, however, closer scrutiny of his work by Leon Kamin (1974; 1981) suggested carelessness and fraud. For instance, Burt was so negligent in his research reports that he even failed to indicate which intelligence tests he used to measure intelligence. Burt also published his papers with co-authors ("Miss Conway") who could never be located. These "co-authors" were totally unknown in the institution listed as their place of employment. They were also unknown to members of the scientific community of that time. They were probably invented by Burt to render his research claims more plausible.
Burt also reported findings of almost identical (and in some cases identical) correlation coefficients for twin samples of gradually increasing size. Since IQ tests are not a precise tool, it is unlikely that an identical IQ will be obtained even when retesting the same person over time. The likelihood of repeatedly obtaining the same or almost identical coefficients while increasing the sample size in Burt's studies is extremely small. Burt also reported that the socioeconomic status of the adoptive households for the twin pairs (supposedly rated by Burt on a six-point scale) was uncorrelated. This would indicate that the twins were distributed rather randomly to families from various socioeconomic strata. This is a methodologically improbable situation: under normal circumstances, separated twins are likely to be placed into at least partly comparable adoptive environments. These methodologically and statistically suspect aspects of Burt's data (see more details in Kamin, 1974; 1981) as well as Burt's reluctance to let other researchers inspect his raw data strongly suggest an uncomfortable conclusion: Burt may have fraudulently manufactured the data to support his belief that intelligence is largely inherited.
Burt's charismatic impact on British social policies adversely affected millions of children in the United Kingdom. According to Fancher, Burt “testified to British government committees that children's intelligence levels were largely fixed by the age of eleven or so, and were accurately measurable by standard tests given at that age. Thus Burt's was one of several influential voices which helped produce the so-called "eleven plus" examination system in Britain, under which all eleven-year-olds were given a series of academic and intelligence tests, the results of which streamed the top-scoring minority into intellectually demanding "grammar schools" and the majority into the less challenging "modern schools." It was virtually impossible for a child to move from a modern to a grammar school, and grammar school training was required for eventual acceptance into a university.” (1985: 176)
Ironically, Burt was promoted to prominent positions within British academic psychology, including chair at the University College of London. As editor of the British Journal of Statistical Psychology, he was able to fill the journal with his own numerous articles, some of which were extremely lengthy and had not the remotest connection with statistical or any other branch of mathematical psychology (Hearnshaw, 1979). His various academic and editorial positions provided him with ample opportunities to selectively promote students with similar hereditarian views: if he were likely to select and promote his protégés on the basis of their personal political views rather than on the basis of their methodological skills, we now may have an aging generation of academic psychologists artificially promoted by his charismatic influence as well as subsequent generations of those similarly promoted by Burt's former protégés. These underlying phenomena could partly explain the contemporary boom of sociobiological or behavior genetics publications in which "heritabilities" are calculated for a wide range of behaviors on the basis of dubious statistical models and unrealistic methodological assumptions (see criticisms presented by Crusio, 1990; Flynn, 1987a; Roubertoux and Capron, 1990a; 1990b; Schonemann, 1989; 1990; 1992; 1995; Taylor, 1980; Wahlsten, 1994).
Thomas Bouchard has been represented by journalists in the last decades as one of the most prominent investigators into heritability of behavior and intelligence. However, both his hereditarian views and his reluctance to submit his twin data from heritability studies to an independent inspection by peer scientists (see Horgan, 1993) are largely reminiscent of Cyril Burt. Without peer scrutiny, research data have very little scientific value.
It is important to emphasize that the static concept of intelligence as primarily genetic in origin conflicts with evidence of a massive increase in IQ scores from one generation to the next. According to Flynn (1987b), data from fourteen economically advanced nations indicate IQ gains ranging from five to twenty-five points in a single generation. This intergenerational increase suggests there are powerful environmental influences that affect performance on tasks typically included in IQ tests. This increase is frequently ignored both by authors of contemporary introductory psychology textbooks and by researchers who study human intellectual potential.
There are other examples of environmental factors in IQ differences. For example, low brain weight found in some old data on blacks from hot African countries could be related to relative infant malnutrition rather than to genetic racial differences. The detrimental impact of malnutrition on brain development has been documented by Monckeberg (1973). Malnutrition affects not only brain and head size but also intelligence. For instance, the quality and quantity of nutritional intake in 153 Egyptian infants aged between eighteen and thirty months was found to be correlated with their intellectual performance at twenty-four months: infants with more adequate nutrition fared better Wachs et al., 1993). Recent studies on adults suggest that even skipping breakfast may result in relatively inferior performance on subsequent cognitive tasks such as those from Cattel's Culture Fair test (Spring et al., 1992).
At present, black parents tend to be younger than white parents at the birth of their first child. Black parents also more frequently have larger families, thus having more later-born children. Statistical reviews by Storfer (1990) show that, on average, children of younger parents have lower IQs than those of older parents and that later-born children from large families have lower IQs than those born first. According to Storfer, these and related factors could explain a large part of IQ discrepancies reported in some comparisons of black and white children.
Unidimensional View of Intelligence
Measuring intelligence in IQ points is reductionistic in the sense that it reduces an individual's intellectual ability to a single number. This is a common practice in clinical, industrial, and school settings. The underlying assumption is that all (or most) facets of intelligence have the same underlying common factor, often labelled "g". The g factor theory has generated lengthy academic debates.
According to Schonemann (1995), the belief in a unitary g factor is irreconcilable with modern statistical methodology. Contemporary mainstream research has not adequately mapped human intelligence. As noted above, this research has mostly been restricted to school-related numerical, verbal, and spatial skills and to the culture-specific information disseminated in public and private schools. To the extent that these tasks require similar skills, reductionist psychologists may find illusory support for the unitary g factor theory, at least in some studies. Traditional IQ tests tend to be the epitome of the school routine. Test takers and test interpreters implicitly share the beliefs that there is only one correct solution for most items (as in convergent thinking), that all viable solutions are known to the test psychologist, and that these correct solutions are only reiterated by the test taker. The reductionist approach fails to include items assessing creativity (as in divergent thinking; see discussion in Cohen et al., 1992). These creative aspects of intelligence are crucial for scientific and economic progress, human adaptability, and the long-term survival of our species.
Greatly stimulating for pioneers exploring neglected aspects of human intelligence is the intuitive monograph by Howard Gardner (1983). Gardner discusses seven intelligences: linguistic, logical-mathematical, spatial, musical, bodily-kinaesthetic (as in dancers, actors, and competitive athletes), intrapersonal (knowing oneself), and interpersonal (knowing others). In sum, defining intelligence as whatever is measured by IQ tests (a definition proposed by some psychologists in the past) unduly restricts the concept of human cognitive functioning.
Reliance on Poor Research: the Examples of Rushton and Lynn
Recent comparisons of intelligence in different racial groups by J. Philippe Rushton (University of Western Ontario, Canada) and Richard Lynn (University of Ulster, Northern Ireland) are remarkable examples of pseudobiological focus (see Lynn, 1993; Rushton, 1988). Unfortunately, their research has had more influence than its quality deserves. In this section I summarise methodological objections to their conclusions about human intelligence.
Head Size Measures as Indicator of Intelligence
Both Rushton and Lynn use head size measures as a convenient substitute for IQ scores, on the inaccurate assumption that these two variables are sufficiently closely related. Rushton (1990a) listed correlation coefficients from twenty studies of head size and intelligence to document the existence of a statistical relationship between the two. The average correlation in Rushton's list (as calculated by Cernovsky, 1991) was only 0.18. This is too low to support Rushton's and Lynn's interpretation of head size differences as differences in intelligence. Classical introductory psychology textbooks warn about similar overinterpretations of low correlation coefficients. For instance, Atkinson et al. state: "Correlations between 0 and .20 must be judged with caution and are only minimally useful in making predictions. One should be suspicious of investigators who make strong claims that are based on correlation coefficients in this lower range" (1983: 24).
As reported in more detail elsewhere (Cernovsky, 1994), Bouchard defended Rushton's reliance on low correlation coefficients. He argued during the question period following Cernovsky's (1992) paper at the International Congress of Psychology in 1992 that weak correlation coefficients should not be underestimated. As support, he referred to Rosenthal's recent work on this issue. But this is a misuse of Rosenthal's work on meta-analysis.
Rosenthal and Rubin (1985) argued that small statistical trends are frequently invaluable even when they fail to reach traditional criteria of statistical significance (p = 0.05 or 0.01). They provide the following example:
“Suppose that, of 20 critically ill patients in a small, randomised experiment, 10 are assigned to a treatment condition and the other 10 are assigned to a control condition. If none of the control patients survive and 3 experimental patients survive, our results will not be significant at p < .05 by a chi-square test or a Fisher exact test. However, we believe it is essential on scientific as well as ethical grounds that such results should be published.” (1985: 528)
In Rosenthal's example, Type I error (use of an ineffective drug to treat a patient who is going to die very soon anyway) has only minor negative consequences compared to Type II error (failure to use treatment that might save 30% of the critically ill patients). The majority of us, if critically ill, would still opt to receive the treatment. That is, we would choose a slight chance of survival even if it is not a "statistically significant" chance, because without the treatment we would die anyway. There are no negative consequences from taking the drug unnecessarily.
This is not at all comparable to claiming, based on weak correlations, that head size is an indicator of intelligence. Type I error, underlying Rushton's speculations, leads not to minor negative consequences but to the defamation of blacks, promotion of racial hatred, and even unnecessary loss of life in racist mob activities. Rosenthal and Rubin defend an experimental use of weak trends in situations where ignoring these trends may result in clearly aversive consequences. But Rushton and Lynn go far beyond this statistical context: they treat two weakly correlated variables as sufficiently identical to rely on the first as the indicator of the second.
Rushton and Lynn also occasionally refer to modern brain size studies. The relationship of brain size to IQ score may be somewhat closer than that found for head size. However, this relationship is also too low to justify substituting one variable for the other. For example, the correlations found by Andreasen et al. (1993) ranged from 0.26 to 0.56, indicating from 12% to 31% of shared variance. Yet, Andreasen and her colleagues emphasized the modest nature of these relationships. Even 31% of shared variance certainly does not justify using one variable as a viable measure of the other. Rushton and Lynn also ignore clinical case studies by Lorber on British adolescents with an extremely small cortex (see summaries in Lewin, 1980). Some of them had IQ scores at or above 120 and were academically successful in high school and subsequently at university, including in areas such as mathematics.
Obsolete Data Sets
Rushton (1988; 1995) and Lynn (1993) frequently rely on antiquated data. As pointed out by Weizmann et al. (1991), old skull collections may have peculiar social histories. For example, both Rushton and Lynn marshalled the skull size data from the famous Morton's collection as evidence of racial inferiority of blacks. Yet the skulls from Morton's collection were originally collected by George Glidden (Stanton, 1965), a supporter of slavery. Glidden may have pre-selected the skulls for each racial group on the basis of skull size in order to support his political position. His motivation was to prove that the creators of ancient Egyptian civilization were white and that blacks existed only in subservient positions.
Misrepresentation of Conclusions of Other Scientists
Both Rushton (1990a; 1990b; 1990c; 1991; 1995) and Lynn (1993) misrepresented statistical analyses by Beals et al. (1984) as supportive of their racial theory. According to Rushton,
“Beals et al. (1984, p.306, Table 2) computerised the entire world database of 20,000 crania gathered by 1940 (after which data collection virtually ceased because of its presumed association with racial prejudice), grouped them by continental area, and found statistically significant differences. Sex-combined brain cases from Asia averaged 1380 cm3 (SD = 83), Europe averaged 1362 cm3 (SD = 35), and Africa averaged 1276 cm3 (SD = 84).” (1990b: 791)
The table with cranial data averages for the continents indeed exists in Beals et al.'s article. However, Rushton and Lynn neglected to mention that Beals et al. explicitly warned readers, on the same page, that these data confound genetic influences with the effects of climatic zone: "If one merely lists such means by geographical region or race, causes of similarity by genogroup and ecotype are hopelessly confounded" (1984: 306).
Within a given racial group, cranial capacity varies depending on the climatic zone. For example, the American Indians are spread over a wide variety of climatic zones and show a corresponding variation in skull size: those from warmer climates have smaller cranial capacity. This pattern is also true for other racial groups. Beals et al. concluded, on the basis of extensive statistical analyses, that correlations of brain size to race are spurious: smaller crania are found in warmer climates, irrespective of race. In fact, Rushton's own tabular summaries of cranial data, based on Herskovits (1930), clearly show these trends. In Rushton's summaries (1990b: see Table 2), the average cranial capacity for North American blacks (1622 cm3) is similar to the average for Caucasians (1621 cm3) from comparable climatic zones. Caucasians from warmer zones such as Cairo (1502 cm3) were similar to some of the black Africans, for example, the Masai (1508 cm3). It is only by "pooling" the black North American data with data for blacks from countries within hot climatic zones (notorious for famine and infant malnutrition that impede brain growth) that Rushton obtained an illusory support for his "genetic" postulates.
Rushton (1988; 1995) also misleads his readers to assume that Tobias's (1970) survey of cranial data supports his theory. Rushton selectively reported only those data from Tobias's monograph that were consistent with his theory. He failed to mention the data sets, also reported by Tobias, showing that cranial size and number of "excess neurones" of North American blacks exceeded those of the French, the English, and American whites (1970: 9, Table 3).
Another example of misrepresenting the work of others: Lynn (1993) reanalysed old data from a study of physical characteristics of Philadelphia school children in the decades preceding 1970, collected by Krogman (1970). He concluded that head sizes are larger in whites than in blacks and also larger in men than in women. According to Lynn, given the positive association between brain size and intelligence, "there should be corresponding race and sex differences in intelligence" (1993: 92). But Lynn misled his readers with respect to the social background of the children in Krogman's study. Allegedly quoting directly from page 4 of Krogman's monograph, Lynn described the blacks as being from the middle and upper-middle class and the white children as being from the middle class. However, on that page Krogman (1970) described the sample of black children as being from the lower-middle and middle-middle class and the white children as being from the middle-middle and upper-middle class. The two racial groups differed with respect to socioeconomic class in the opposite direction than sketched by Lynn. According to Krogman, the whites were largely from the white collar and skilled labour population with a few from a professional and academic level. The blacks were from the blue collar and semi-skilled labour population. Since the data were collected in the decades preceding 1970, these class differences in urban settings could well be associated with a major difference in quality of child health care, nutrition, educational stimulation by parents, and other factors with a potentially adverse impact on the central nervous system and intellectual development.
Several types of brain size measures can be used when comparing samples grouped on the basis of gender or skin colour. Some researchers consider it important to correct the absolute brain size for body height, weight, or surface, on the assumption that larger bodies require proportionately larger brains for the control of various motor and physical functions. Rushton's and Lynn's uncritical interpretation of these measures of brain/body size ratio as a valid indicator of biological intelligence is misleading: some lower animals such as squirrel monkeys or house mice have more favourable brain/body ratios than humans without demonstrating a corresponding intellectual superiority (see a review of Rushton's theory by Cain and Vanderwolf 1990). Both Rushton and Lynn almost always rely only on absolute brain (or cranial) size data. Significantly, they resort to corrections of cranial size data for body size only when convenient to defend their dogma of black inferiority. For example, Lynn (1993) built almost his entire article (and his thesis of blacks' intellectual inferiority) exclusively on the review of absolute cranial size data. Yet he then suddenly switched to the cranial size "corrected" for body height when the absolute cranial capacity in black girls was greater than in white girls.
Meta-Analysis of Rushton 's Data by Gorey and Cryns
A recent meta-analytic study provides additional evidence of methodological weakness in Rushton's procedures. Gorey and Cryns (1995) indicated that the mean correlation coefficient in data sets listed by Rushton for black-white intelligence differences was only 0.23. Gorey and Cryns also found that Rushton's work has an unbalanced overrepresentation of references to supportive data. When they recalculated the data for the same variables based on a computerised random literature search, the mean correlation coefficient dropped to 0.15. Similarly, the mean coefficient based on data sets chosen by Rushton to document black-white differences in personality and temperament was 0.37. But with data based on a random literature search, this coefficient dropped to -0.02. Gorey and Cryns examined eight variables. On all eight, Rushton's "data" were more supportive of his hypotheses than the data based on a computerised random literature search.
Preferential Publication of Supportive Data
Some researchers (and also journal editors) may hesitate to publish data not supportive of their personal beliefs. Or, while gathering the data, their data can somehow be distorted to match personal expectations. For instance, Rosenthal's (1991) statistical work suggested that about two-thirds of observational errors made by investigators are in the direction of supporting their hypothesis. These biased errors occasionally push a result over the magic 0.05 cliff, leading to "statistically significant" confirmatory findings.
In some cases, political bias or political pressure may lead to a preponderance of published findings that are unfavourable to blacks. For example, the editor of Canadian Psychology, the leading Canadian psychological journal, published a lengthy article by Rushton defending his racial research and expounding his racial theory. Subsequently, the editor declined to publish a manuscript describing methodological and statistical deficiencies in Rushton's work. He said that "each new salvo against Dr. Rushton's position inevitably requires that he be permitted the right of response, and his views - which do not seem to change - are trotted out again and again" (letter from Patrick O'Neill, editor, Canadian Psychology, June 20, 1991). The editor's letter admitted that the submitted criticism of Rushton may present original methodological information. Yet, these methodological considerations were less relevant for the editorial decision than "academic politics" in Canadian circles.
Until recently, black students were systematically prevented (often by the brutality of the white mob) from entering US universities. Some who persisted paid for this with their lives. For this reason, most data are from studies prepared by white psychologists only, some of whom are notorious for their bias against blacks. Their biased research style is exemplified by Arthur Jensen's work. His findings of racial differences in reaction time to complex tasks (interpreted as a relatively culture-free measure of intellectual functioning) have been artificially manufactured by his selective publishing of confirmatory results only. He has failed to publish his own disconfirmatory data (see documentation by Kamin and Grant-Henry, 1987).
The critical psychologist may deal with a wide variety of topics in intelligence research. The issues raised here partly overlap those discussed by Kamin (1995) in his critique of the highly publicised controversial book The Bell Curve (Herrnstein and Murray, 1994). Herrnstein and Murray erroneously treat Lynn's and Rushton's work as a reliable scientific source. They fail to discuss the numerous methodological flaws in Rushton's and Lynn's methodology as known from various reviews (Cain and Vanderwolf, 1990; Flynn, 1989; 1990; Kamin, 1995; Weizmann et al., 1990; 1991; Zuckerman and Brody, 1988).
The widespread promotion of the works of Rushton and Lynn and recent attempts to rescue the scientific reputation of Cyril Burt (see Jensen, 1992) are alarming academic developments. These trends may discredit American and Canadian academic psychology on an international scale. Research funding could be channelled more constructively into more methodologically adequate research on personality factors and on strategies that enhance economic or scientific creativity and productivity.
Anastasi, A. (1988). Psychological Testing (6th edn). New York: Macmillan.
Andreasen, N., Flaum, M., Swayze, V., O'Leary, D. S., Alliger, R., Cohen, G., Ehrhardt, J., and Yuh, W. T. C. (1993). Intelligence and Brain Structure in Normal Individuals. American Journal of Psychiatry, 150, 130-134.
Atkinson, R. L., Atkinson, R. C., and Hilgard, E. R. (1983). Introduction to Psychology (8th edn). New York: Harcourt Brace Jovanovich.
Beals, K. L., Smith, C. L., and Dodd, S. M. (1984). Brain Size, Cranial Morphology, Climate, and Time Machines. Current Anthropology, 25, 301-330.
Cain, D. P., and Vanderwolf, C. H. (1990). A Critique of Rushton on Race, Brain Size, and Intelligence. Personality and Individual Differences, 11, 777-784.
Cernovsky, Z. Z. (1991). Intelligence and Race: Further Comments on J.P. Rushton's Work. Psychological Reports, 68, 481-482.
Cernovsky, Z. Z. (1992). J.P. Rushton's Aggregational Errors in Racial Psychology. Paper presented at the 25th International Congress of Psychology, Brussels, Belgium, July.
Cernovsky, z. z. (1994). Rushton's Defenders and their Hasty Rejection of the Null Hypothesis. Journal of Black Psychology, 20, 325-333.
Cohen, R. J., Swerdlik, M. E., and Smith, D. K. (1992). Psychological Testing and Assessment. Mountain View, CA: Mayfield.
Crusio, W. E. (1990). Intelligent Quantitative Genetics: Asking the Right Questions. Cahiers de Psychologie Cognitive, 10, 619-25.
Evans, B., and Waites, B. (1981). IQ and Mental Testing. London: Macmillan.
Fancher, R. E. (1985). The Intelligence Men: Makers of the IQ Controversy. New York and London: Norton.
Fancher, R. E. (1988). Henry Goddard and the Kallikak Family Photographs: "Conscious Skulduggery" or "Whig History"? American Psychologist, 42, 585-590.
Flynn, J. R. (1987a). Race and IQ: Jensen's Case Refuted. In S. Modgil and C. Modgil (Eds.), Arthur Jensen: Consensus and Controversy (pp. 221-232). New York: Falmer Press.
Flynn, J. R. (1987b). Massive IQ Gains in 14 Nations: What IQ Tests Really Measure. Psychological Bulletin, 101,171-191.
Flynn, J. R. (1989). Rushton, Evolution, and Race: An Essay on Intelligence and Virtue. The Psychologist Bulletin of the British Psychological Society, 9, 363-366.
Flynn, J. R. (1990). Explanation, Evaluation, and a Rejoinder to Rushton. The Psychologist Bulletin of the British Psychological Society, 5, 199-200.
Gardner, H. (1983). Frames of Mind The Theory of Multiple Intelligences. New York: Basic Books.
Goddard, H. H. (1917). Mental Tests and the Immigrant. Journal of Delinquency, 2, 243-277.
Gorey, K. M., and Cryns, A. G. (1995). Lack of Racial Differences in Behavior: A Quantitative Replication of Rushton's (1988) Review and an Independent Meta-Analysis. Personality and Individual Differences, 19, 345-353.
Hearnshaw, L. 5. (1979). Cyril Burt Psychologist London: Hodder and Stoughton.
Herrnstein, R. J., and Murray, C. (1994). The Bell Curve: Intelligence and Class Structure in American Life. New York: Free Press.
Herskovits, M. J. (1930). The Anthropometry of the American Negro. New York: Columbia University Press.
Horgan, J. (1993). Eugenics Revisited. Scientific American, 268(6), 123-131.
Jensen, A. R. (1969). How Much Can We Boost IQ and Scholastic Achievement? Harvard Educational Review, 39, 1-123.
Jensen, A. R. (1992). Scientific Fraud or False Accusations? The Case of Cyril Burt. In D.J. Miller and M. Hersen (Eds.), Research Fraud in the Behavioral and Biomedical Sciences (pp. 97-124). New York: Wiley.
Kamin, L. (1974). The Science and Politics of IQ. Potomac, MD: Erlbaum.
Kamin, L. (1981). Chapters 12 to 20, and 22, in H. J. Eysenck and L. Kamin (Eds.), Intelligence: The Battle for the Mind. H. J Eysenck versus Leon Kamin. London: Macmillan.
Kamin, L. (1995). Behind the Curve. Scientific American, 272, 99-103.
Kamin, L., and Grant-Henry, S. (1987). Reaction Time, Race, and Racism. Intelligence, 11, 299-304.
Krogman, W. M. (1970). Growth of Head, Face, Trunk, and Limbs in Philadelphia White and Negro Children of Elementary and High School Age. Monographs of the Society of Research on Child Development, 35, no. 136.
Lewin, R. (1980). Is Your Brain Really Necessary? Science, 210, 1232-1234.
Lynn, R. (1993) Further Evidence for the Existence of Race and Sex Differences in Cranial Capacity. Social Behavior and Personality, 21, 89-92.
Miller-Jones, D. (1989). Culture and Testing. American Psychologist, 44, 360-366.
Monckeberg, F. B. (1973). Effects of Nutrition on Brain and Intellectual Development. In F. Richardson (Ed.), Brain and Intelligence (pp. 207-236). Hyattsville, MD: National Educational Press.
Rosenthal, R. (1991). Meta-Analytic Procedures for Social Sciences. London: Sage.
Rosenthal, R., and Rubin, D. B. (1985). Statistical Analysis: Summarising Evidence versus Establishing Facts. Psychological Bulletin, 97, 527-529.
Roubertoux, P. L., and Capron, C. (1990a). Are Intelligence Differences Hereditarily Transmitted? Cahiers de Psychologie Cognitive, 10, 555-594.
Roubertoux, P. L., and Capron, C. (1990b). Now to the Future: The Heritability of IQ versus the Cognitive-Genetic Analysis. Cahiers de Psychologie Cognitive, 10, 715-721.
Rushton, J. P. (1988). Race Differences in Behavior: A Review and Evolutionary Analysis. Personality and Individual Differences, 9, 1009-1024.
Rushton, J. P. (1990a). Race, Brain Size, and Intelligence: A Reply to Cernovsky. Psychological Reports, 66, 659-666.
Rushton, J. P. (1990b). Race, Brain Size, and Intelligence: A Rejoinder to Cain and Vanderwolf. Personality and Individual Differences, 11, 785-794.
Rushton, J. P. (1990c). Race and Crime: Reply to Roberts and Gabor. Canadian Journal of Criminology, 32, 315-334.
Rushton, J. P. (1991). Do r/k Strategies Underlie Human Race Differences? A Reply to Weizmann et al. Canadian Psychology, 32, 29-42.
Rushton, J. P. (1995). Race, Evolution, and Behavior. Brunswick, NJ: Transaction.
Schonemann, P. (1989). Some New Results on the Spearman Hypothesis Artefact. Bulletin of the Psychonomic Society, 27, 462-464.
Schonemann, P. (1990). Environmental versus Genetic Variance Component Models for Identical Twins: A Critique of Jink and Fulker's Reanalysis of the Shields Data. Cahiers de Psychologie Cognitive, 10, 451-473.
Schonemann, P. (1992). Extension of Guttman’s Result from g to PCI. Multivariate Behavioral Research, 27(2), 219-224.
Schonemann, P. (1995). Totems of the IQ Myth and its Heritabilities: g and IC. Paper presented at Meeting of the American Association for the Advancement of Science, Atlanta, Georgia, February.
Spring, B., Pingitore, R., Bourgeois, M., Kessler, K. H., and Bruckner, E. (1992). The Effects and Non-Effects of Skipping Breakfast: Results of Three Studies. Paper presented at the 100th Annual Meeting of the American Psychological Association, Washington, DC, August.
Stanton, W. (1965). The Leopard's Spots. Chicago: University of Chicago Press.
Storfer, M. D. (1990). Intelligence and giftedness. San Francisco: Jossey-Bass.
Taylor, H. F. (1980). The IQ Game: A Methodological Inquiry into the Heredity-Environment Controversy. New Brunswick, NJ: Rutgers University Press.
Terman, L. (1919). The Measurement of Intelligence. London: Harrap.
Tobias, P. V. (1970). Brain Size, Grey Matter, and Race: Fact or Fiction? American Journal of Physical Anthropology, 32, 3-25.
Wachs, T. D., Moussa, W., Bishry, Z., Yunis, F., Sobhy, A., McCabe, G., Jerome, N., Galal, O., Harrison, G., and Kirksey, A. (1993). Relations between Nutrition and Cognitive Performance in Egyptian Toddlers. Intelligence, 17, 151-172.
Wahlsten, D. (1994). The Intelligence of Heritability. Canadian Journal of Psychology, 35, 244-260.
Weizmann, F., Wiener, N. I., Wiesenthal, D. L., and Ziegler, M. (1990). Differential K Theory and Racial Hierarchies. Canadian Psychology, 31, 1-13.
Weizmann, F., Wiener, N. I., Wiesenthal, D. L., and Ziegler, M. (1991). Eggs, Eggplants, and Eggheads: A Rejoinder to Rushton. Canadian Psychology, 32, 43-50.
Zuckerman, M., and Brody, N. (1988). Oysters, Rabbits, and People: A Critique of "Race Differences in Behavior" by J. P. Rushton. Personality and Individual Differences, 9, 1025-1033.