An Intelligence Quotient or IQ is a score derived from one of several different standardized tests attempting to measure intelligence. The term "IQ," a translation of the German Intelligenz-Quotient, was coined by the German psychologist William Stern in 1912 as a proposed method of scoring early modern children's intelligence tests such as those developed by Alfred Binet and Theodore Simon in the early 20th Century. Although the term "IQ" is still in common use, the scoring of modern IQ tests such as the Wechsler Adult Intelligence Scale is now based on a projection of the subject's measured rank on the Gaussian bell curve with a center value (average IQ) of 100, and a standard deviation of 15 (different tests have various standard deviations; the Stanford-Binet IQ test has a standard deviation of 16).
IQ scores have been shown to correlate with such factors as morbidity and mortality, parental social status, and to a substantial degree, parental IQ. While its inheritance has been investigated for nearly a century, controversy remains as to how much is inheritable, and the mechanisms of inheritance are still a matter of some debate.
IQ scores are used in many contexts: as predictors of educational achievement or special needs, by social scientists who study the distribution of IQ scores in populations and the relationships between IQ score and other variables, and as predictors of job performance and income.
The average IQ scores for many populations have been rising at an average rate of three points per decade since the early 20th century with most of the increase in the lower half of the IQ range: a phenomenon called the Flynn effect. It is disputed whether these changes in scores reflect real changes in intellectual abilities, or merely methodological problems with past or present testing.
In 1912, the German psychologist William Stern coined the abbreviation "I.Q.," a translation of the German Intelligenz-Quotient ("intelligence quotient"), proposing that an individual's intelligence level be measured as a quotient of their estimated "mental age" and their chronological age. A further refinement of the Binet-Simon scale was published in 1916 by Lewis M. Terman, from Stanford University, who incorporated Stern's proposal, and this Stanford-Binet Intelligence Scale formed the basis for one of the modern intelligence tests that remains in common use.
At first, IQ was calculated as a ratio with the formula
In 1939 David Wechsler published the first intelligence test explicitly designed for an adult population, the Wechsler Adult Intelligence Scale, or WAIS. Subsequent to the publication of the WAIS, Wechsler extended his scale for younger ages, creating the Wechsler Intelligence Scale for Children, or WISC. The Wechsler scales contained separate subscores for verbal and performance IQ, thus being less dependent on overall verbal ability than early versions of the Stanford-Binet scale, and was the first intelligence scale to base scores on a standardized normal distribution rather than an age-based quotient: since age-based quotients worked only for children, these methods were replaced by a projection of the measured rank on the Gaussian bell curve using an average IQ of 100 as the center value and a standard deviation of 15 or occasionally 16 or 24 points.
Thus, the modern IQ score is a mathematical transformation of a raw score on an IQ test, based on the rank of that score in a normalization sample, Modern scores are sometimes referred to as "deviance IQ", while older method age-specific scores are referred to as "ratio IQ."
The two methodologies yield similar results near the middle of the bell curve, but the older ratio IQs yielded far higher scores for the intellectually gifted— for example, Marilyn vos Savant, who appeared in the Guinness Book of World Records, obtained a ratio IQ of 228. While this score could make sense using Binet's formula (and even then, only for a child), on the Gaussian curve model it would be an exceptional 7.9 standard deviations above the mean and hence virtually impossible in a population with a normal IQ distribution (see normal distribution). In addition, IQ tests like the Wechsler were not intended to discriminate reliably much beyond IQ 130, as they simply do not contain enough exceptionally difficult items.
Since the publication of the WAIS, almost all intelligence scales have adopted the normal distribution method of scoring. The use of the normal distribution scoring method makes the term "intelligence quotient" an inaccurate description of the intelligence measurement, but "I.Q." still enjoys colloquial usage, and is used to describe all of the intelligence scales currently in use.
The effect of restriction of range on IQ was examined by Matt McGue and colleagues, who wrote that "restriction in range in parent disinhibitory psychopathology and family SES had no effect on adoptive-sibling correlations ... IQ. On the other hand, a 2003 study by Eric Turkheimer, Andreana Haley, Mary Waldron, Brian D'Onofrio, Irving I. Gottesman demonstrated that the proportions of IQ variance attributable to genes and environment vary with socioeconomic status. They found that in impoverished families, 60% of the variance in IQ "in a sample of 7-year-old twins" is accounted for by the shared environment, and the contribution of genes was close to zero.
It is reasonable to expect that genetic influences on traits like IQ should become less important as one gains experiences with age. Surprisingly, the opposite occurs. Heritability measures in infancy are as low as 20%, around 40% in middle childhood, and as high as 80% in adulthood. The American Psychological Association's 1995 task force on "Intelligence: Knowns and Unknowns" concluded that within the white population the heritability of IQ is "around .75." The Minnesota Study of Twins Reared Apart, a multiyear study of 100 sets of reared-apart twins which was started in 1979, concluded that about 70% of the variance in IQ was found to be associated with genetic variation. Some of the correlation of IQs of twins may be a result of the effect of the maternal environment before birth, shedding some light on why IQ correlation between twins reared apart is so robust.
There are a number of points to consider when interpreting heritability:
A study of French children adopted between the ages of 4 and 6 shows a possible continuing interplay of nature and nurture. The children came from poor backgrounds with IQs that initially averaged 77, putting them near retardation. Nine years later after adoption, they retook the I.Q. tests, and all of them did better. The amount they improved was directly related to the adopting family’s status. "Children adopted by farmers and laborers had average I.Q. scores of 85.5; those placed with middle-class families had average scores of 92. The average I.Q. scores of youngsters placed in well-to-do homes climbed more than 20 points, to 98. On the other hand, the degree to which these increases persisted into adulthood are not clear from the study.
Eric Turkheimer and colleagues (2003), not using an adoption study, included impoverished US families. Results demonstrated that the proportions of IQ variance attributable to genes and environment vary nonlinearly with socio-economic status. The models suggest that in impoverished families, 60% of the variance in IQ is accounted for by the shared family environment, and the contribution of genes is close to zero; in affluent families, the result is almost exactly the reverse. They suggest that the role of shared environmental factors may have been underestimated in older studies which often only studied affluent middle class families.
A meta-analysis, by Devlin and colleagues in Nature (1997), of 212 previous studies evaluated an alternative model for environmental influence and found that it fits the data better than the 'family-environments' model commonly used. The shared maternal (fetal) environment effects, often assumed to be negligible, account for 20% of covariance between twins and 5% between siblings, and the effects of genes are correspondingly reduced, with two measures of heritability being less than 50%.
Bouchard and McGue reviewed the literature in 2003, arguing that Devlin's conclusions about the magnitude of heritability is not substantially different than previous reports and that their conclusions regarding prenatal effects stands in contradiction to many previous reports. They write that:
Chipuer et al. and Loehlin conclude that the postnatal rather than the prenatal environment is most important. The Devlin et al conclusion that the prenatal environment contributes to twin IQ similarity is especially remarkable given the existence of an extensive empirical literature on prenatal effects. Price (1950), in a comprehensive review published over 50 years ago, argued that almost all MZ twin prenatal effects produced differences rather than similarities. As of 1950 the literature on the topic was so large that the entire bibliography was not published. It was finally published in 1978 with an additional 260 references. At that time Price reiterated his earlier conclusion. Research subsequent to the 1978 review largely reinforces Price’s hypothesis.
In 2004, Richard Haier, professor of psychology in the Department of Pediatrics and colleagues at University of California, Irvine and the University of New Mexico used MRI to obtain structural images of the brain in 47 normal adults who also took standard IQ tests. The study demonstrated that general human intelligence appears to be based on the volume and location of gray matter tissue in the brain. Regional distribution of gray matter in humans is highly heritable. The study also demonstrated that, of the brain's gray matter, only about 6 percent appeared to be related to IQ.
Many different sources of information have converged on the view that the frontal lobes are critical for fluid intelligence. Patients with damage to the frontal lobe are impaired on fluid intelligence tests (Duncan et al 1995). The volume of frontal grey (Thompson et al 2001)
Much research has been devoted to the extent and potential causes of racial group differences in IQ.
While IQ is sometimes treated as an end unto itself, scholarly work on IQ focuses to a large extent on IQ's validity, that is, the degree to which IQ correlates with outcomes such as job performance, social pathologies, or academic achievement. Different IQ tests differ in their validity for various outcomes. Traditionally, correlation for IQ and outcomes is viewed as a means to also predict performance; however readers should distinguish between prediction in the hard sciences and the social sciences.
Validity is the correlation between score (in this case cognitive ability, as measured, typically, by a paper-and-pencil test) and outcome (in this case job performance, as measured by a range of factors including supervisor ratings, promotions, training success, and tenure), and ranges between −1.0 (the score is perfectly wrong in predicting outcome) and 1.0 (the score perfectly predicts the outcome). See validity (psychometric).
Research shows that general intelligence plays an important role in many valued life outcomes. In addition to academic success, IQ correlates to some degree with job performance (see below), socioeconomic advancement (e.g., level of education, occupation, and income), and "social pathology" (e.g., adult criminality, poverty, unemployment, dependence on welfare, children outside of marriage). Recent work has demonstrated links between general intelligence and health, longevity, and functional literacy. Correlations between g and life outcomes are pervasive, though IQ does not correlate with subjective self-reports of happiness. IQ and g correlate highly with school performance and job performance, less so with occupational prestige, moderately with income, and to a small degree with law-abiding behaviour. IQ does not explain the inheritance of economic status and wealth.
Correlations between IQ scores (general cognitive ability) and achievement test scores are reported to be .81 by Deary and colleagues, with the percentage of variance accounted for by general cognitive ability ranging "from 58.6% in Mathematics and 48% in English to 18.1% in Art and Design".
Correlations between IQ scores and total years of education are about .55, implying that differences in psychometric intelligence account for about 30% of the outcome variance. Many occupations can only be entered through professional schools which base their admissions at least partly on test scores: the MCAT, the GMAT, the GRE, the DAT, the LSAT, etc. Individual scores on admission-related tests such as these are certainly correlated with scores on tests of intelligence. It is partly because intelligence test scores predict years of education that they also predict occupational status, and income to a smaller extent.
According to Schmidt and Hunter, "for hiring employees without previous experience in the job the most valid predictor of future performance is general mental ability." The validity depends on the type of job and varies across different studies, ranging from 0.2 to 0.6. However IQ mostly correlates with cognitive ability only if IQ scores are below average and this rule has many (about 30 %) exceptions for people with average and higher IQ scores. Also, IQ is related to the "academic tasks" (auditory and linguistic measures, memory tasks, academic achievement levels) and much less related to tasks where even precise hand work ("motor functions") are required.
A meta-analysis which pooled validity results across many studies encompassing thousands of workers (32,124 for cognitive ability), reports that the validity of cognitive ability for entry-level jobs is 0.54, larger than any other measure including job try-out (0.44), experience (0.18), interview (0.14), age (−0.01), education (0.10), and biographical inventory (0.37). This implies that, across a wide range of occupations, intelligence test performance accounts for some 29% of the variance in job performance.
According to Marley Watkins and colleagues, IQ is a causal influence on future academic achievement, whereas academic achievement does not substantially influence future IQ scores. Treena Eileen Rohde and Lee Anne Thompson write that general cognitive ability but not specific ability scores predict academic achievement, with the exception that processing speed and spatial ability predict performance on the SAT math beyond the effect of general cognitive ability.
The American Psychological Association's report Intelligence: Knowns and Unknowns states that other individual characteristics such as interpersonal skills, aspects of personality, et cetera, are probably of equal or greater importance, but at this point we do not have equally reliable instruments to measure them.
Other studies show that ability and performance for jobs are linearly related, such that at all IQ levels, an increase in IQ translates into a concomitant increase in performance. Charles Murray, coauthor of The Bell Curve, found that IQ has a substantial effect on income independently of family background.
The American Psychological Association's report Intelligence: Knowns and Unknowns states that IQ scores account for about one-fourth of the social status variance and one-sixth of the income variance. Statistical controls for parental SES eliminate about a quarter of this predictive power. Psychometric intelligence appears as only one of a great many factors that influence social outcomes.
One reason why some studies claim that IQ only accounts for a sixth of the variation in income is because many studies are based on young adults (many of whom have not yet completed their education). On pg 568 of The g Factor, Arthur Jensen claims that although the correlation between IQ and income averages a moderate 0.4 (one sixth or 16% of the variance), the relationship increases with age, and peaks at middle age when people have reached their maximum career potential. In the book, A Question of Intelligence, Daniel Seligman cites an IQ income correlation of 0.5 (25% of the variance).
A 2002 study further examined the impact of non-IQ factors on income and concluded that an offspring's inherited wealth, race, and schooling are more important as factors in determining income than IQ.
There is a correlation of -0.19 between IQ scores and number of juvenile offences in a large Danish sample; with social class controlled, the correlation dropped to -0.17. Similarly, the correlations for most "negative outcome" variables are typically smaller than 0.20, which means that test scores are associated with less than 4% of their total variance. It is important to realize that the causal links between psychometric ability and social outcomes may be indirect. Children with poor scholastic performance may feel alienated. Consequently, they may be more likely to engage in delinquent behavior, compared to other children who do well.
IQ is also negatively correlated with certain diseases.
Tambs et al. found that occupational status, educational attainment, and IQ are individually heritable; and further found that "genetic variance influencing educational attainment ... contributed approximately one-fourth of the genetic variance for occupational status and nearly half the genetic variance for IQ". In a sample of U.S. siblings, Rowe et al. report that the inequality in education and income was predominantly due to genes, with shared environmental factors playing a subordinate role.
In the United States, certain public policies and laws regarding military service, education, public benefits, crime, and employment incorporate an individual's IQ or similar measurements into their decisions. However, in 1971, for the purpose of minimizing employment practices that disparately impacted racial minorities, the U.S. Supreme Court banned the use of IQ tests in employment, except in very rare cases. Internationally, certain public policies, such as improving nutrition and prohibiting neurotoxins, have as one of their goals raising or preventing a decline in intelligence.
Binet had designed the Binet-Simon intelligence scale in order to identify students who needed special help in coping with the school curriculum. He argued that with proper remedial education programs, most students regardless of background could catch up and perform quite well in school. He did not believe that intelligence was a measurable fixed entity.
He spent much of the book criticizing the concept of IQ, including a historical discussion of how the IQ tests were created and a technical discussion of why g is simply a mathematical artifact. Later editions of the book included criticism of The Bell Curve.
Gould did not dispute the stability of test scores, nor the fact that they predict certain forms of achievement. He did argue, however, that to base a concept of intelligence on these test scores alone is to ignore many important aspects of mental ability.
However, IQ tests may well be biased when used in other situations. A 2005 study stated that "differential validity in prediction suggests that the WAIS-R test may contain cultural influences that reduce the validity of the WAIS-R as a measure of cognitive ability for Mexican American students, indicating a weaker positive correlation relative to sampled white students. Other recent studies have questioned the culture-fairness of IQ tests when used in South Africa. Standard intelligence tests, such as the Stanford-Binet, are often inappropriate for children with autism; and dyslexia the alternative of using developmental or adaptive skills measures are relatively poor measures of intelligence in autistic children, and have resulted in incorrect claims that a majority of children with autism are mentally retarded.
Some argue that IQ scores are used as an excuse for not trying to reduce poverty or otherwise improve living standards for all. Claimed low intelligence has historically been used to justify the feudal system and unequal treatment of women (but note that many studies find identical average IQs among men and women; see sex and intelligence). In contrast, others claim that the refusal of "high-IQ elites" to take IQ seriously as a cause of inequality is itself immoral.
In response to the controversy surrounding The Bell Curve, the American Psychological Association's Board of Scientific Affairs established a task force in 1995 to write a consensus statement on the state of intelligence research which could be used by all sides as a basis for discussion. The full text of the report is available through several websites.
In this paper the representatives of the association regret that IQ-related works are frequently written with a view to their political consequences: "research findings were often assessed not so much on their merits or their scientific standing as on their supposed political implications".
The task force concluded that IQ scores do have high predictive validity for individual differences in school achievement. They confirm the predictive validity of IQ for adult occupational status, even when variables such as education and family background have been statistically controlled. They agree that individual differences in intelligence are substantially influenced by genetics and that both genes and environment, in complex interplay, are essential to the development of intellectual competence.
They state there is little evidence to show that childhood diet influences intelligence except in cases of severe malnutrition. The task force agrees that large differences do exist between the average IQ scores of blacks and whites, and that these differences cannot be attributed to biases in test construction. The task force suggests that explanations based on social status and cultural differences are possible, and that environmental factors have raised mean test scores in many populations. Regarding genetic causes, they noted that there is not much direct evidence on this point, but what little there is fails to support the genetic hypothesis.
The APA journal that published the statement, American Psychologist, subsequently published eleven critical responses in January 1997, several of them arguing that the report failed to examine adequately the evidence for partly-genetic explanations.