Intelligence Quotient

An intelligence quotient or IQ is a score derived from a set of standardized tests that were developed with the purpose of measuring a person's cognitive abilities ("intelligence") in relation to their age group. IQ tests do not measure intelligence the way a ruler measures height, but rather the way a race measures speed. Also, IQ tests measure actual performance, not innate potential. Among the middle class of industrial societies, IQ is highly heritable, and by adulthood the influence of family environment on IQ is undetectable. IQ test scores are correlated with measures of brain structure and function, as well as performance on simple tasks that anyone can complete within a few seconds. IQ is strongly correlated with academic success, but can also predict important life outcomes such as job performance, socioeconomic advancement, and "social pathologies". Recent work has demonstrated links between IQ and health, longevity, and functional literacy.

History

The modern field of IQ testing started with the Stanford-Binet test. Alfred Binet and his colleague Theodore Simon created the IQ test in 1905, aimed at identifying students who could benefit from extra help in school. Their assumption was that a lower IQ indicated the need for more teaching, not an inability to learn. This interpretation is still held by some modern experts. The term "intelligence quotient" comes from Binet's test, in which each student's score was the quotient of his or her tested academic age with his or her actual age.

Today, the most commonly administered IQ test is the WISC-III test, originally developed by David Wechsler in 1974. The WISC-III test comprises ten types of problems, categorized by difficulty and by skill type (verbal and performance scales). Another notable type of IQ test is the Bailey Scale of Infant Development, regarded as the 'best' means of testing cognitive development in infants.

Online Tests

Although online IQ tests have become wildly popular with the explosion of the internet in recent years, they are highly inaccurate. Comparing results among a large set of people shows a common factor: most scores are above 110. Of course, such tests automatically measure very few people in the 70 to 90 range, and hence create a strong upward distortion. Many of these websites do not show the results immediately and instead attempt to sell certificates showing the results.

Distribution

IQ scores are expressed as a number normalized so that the average IQ in an age group is 100 — in other words an individual scoring 115 is above-average when compared to similarly aged people. It is common, but not invariable, practice to standardize so that the standard deviation (s) of scores is 15 or 16 ( 24 is used occasionally as well ). Tests are designed so that the distribution of IQ scores is Gaussian, that is to say that it follows a bell curve.

(The following numbers apply to IQ scales standard deviation s = 15.) Roughly 68% of the population has an IQ between 85 and 115. The "normal" range, or range between -2 and +2 standard deviations from the mean, is between 70 and 130, and contains about 95% of the population. A score below 70 may indicate mental retardation, and a score above 130 may indicate intellectual giftedness. Retardation may result from normal variation or from a genetic or developmental malady; analogously, some otherwise normal people are very short, and others have dwarfism. Giftedness appears to be normal variation; autistic savants have often astonishing cognitive powers but below-average IQs.

It has been observed that scores outside the range 55 to 145 must be cautiously interpreted because there are smaller numbers of respondents with which to make comparisons in those ranges. Moreover, at such extreme values, the normal distribution is a less accurate estimate of the true IQ distribution.

In actuality there is a higher percentage of the population measured at 3sd+ levels on the test than the probabilities of the normal distribution would predict. Some IQ scoring procedures may attempt to integrate such clusters of statistical outliers into the curve by adjusting the scores so that they better represent actual probabilities (according to Silverman) and in these cases, scores around 145 and above may actually have been notably higher, were they not so adjusted.

Children's test scores are based on a simple ratio of a child's "mental age" (representing the age at which the subject's test performance would normally be expected) divided by the child's chronological age, times 100. (MA/CA) x 100 = IQ By age 14 and above, youth and adults are typically tested using adult-normed measures that no longer use simple ratios for IQ determination.

IQ and g

Modern IQ tests produce scores for different areas (e.g., language fluency, three-dimensional thinking, etc.), with the summary score calculated from subtest scores. Individual subtest scores tend to correlate with one another, even when seemingly disparate in content. Analyses of an individual's scores on a wide variety of tests (Stanford-Binet, WISC-R, Raven's Progressive Matrices and others) will reveal that they all measure a single common factor and various factors that are specific to each test. This kind of factor analysis has led to the theory that underlying these disparate cognitive tasks is a single factor, termed the General_intelligence_factor, that represents the common-sense concept of intelligence. In the normal population, g and IQ are roughly 90% correlated and are often used interchangeably.

Genetics vs environment

The role of genes and environment in determining IQ is reviewed in Plomin et al. (2001, 2003). The degree to which genetic variation contributes to observed variation in a trait is measured by a statistic called heritability. Heritability scores range from 0 to 1, and can be interpreted as the percentage of variation (e.g. in IQ) that is due to variation in genes. Twins studies and adoption studies are commonly used to determine the heritability of a trait. Until recently heritability was mostly studied in children. These studies yield an estimate of heritability of 0.5; that is, half of the variation in IQ among the children studied was due to variation in their genes. The remaining half was thus due to environmental variation and measurement error. A heritability of 0.5 implies that IQ is "substantially" heritable.

Considerable research has focused on biological correlates of g; see General_intelligence_factor and the section on brain size below. For example, general intelligence and MRI brain volume measurements are correlated, and the effect is primarily determined by genetic factors.

Environment

Nearly all personality traits show that, contrary to expectations, environmental effects actually cause adoptive siblings raised in the same family to be as different as children raised in different families (Harris, 1998; Plomin & Daniels, 1987). Put another way, shared environmental variation for personality is zero, and all environmental effects are nonshared. Intelligence is actually an exception to this rule, at least among children. The IQs of adoptive siblings, who share no genetic relation but do share a common family environment, are correlated at .32. Despite attempts to isolate them, the factors that cause adoptive siblings to be similar have not been identified. However, as explained below, shared family effects on IQ disappear after adolescence.

Active genotype-environment correlation, also called the "nature of nurture", is observed for IQ. This phenomenon is measured similarly to heritability; but instead of measuring variation in IQ due to genes, variation in environment due to genes is determined. One study found that 40% of variation in measures of home environment are accounted for by genetic variation. This suggests that the way human beings craft their environment is due in part to genetic influences.

Environmental factors may play a larger role in determining IQ in situations where environmental conditions are more variable. Proper childhood nutrition appears critical for cognitive development; malnutrition can lower IQ. Other research indicates environmental factors such as prenatal exposure to toxins, duration of breastfeeding, and micronutrient deficiency can affect IQ.

Development

It is reasonable to expect that genetic influences on traits like IQ should become less important as we gain 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.

Shared family effects also seem to disappear by adulthood. Adoption studies show that, after adolescence, adopted siblings are no more similar in IQ than strangers (IQ correlation near zero), while full siblings show an IQ correlation of 0.6. Twin studies reinforce this pattern: monozygotic (identical) twins raised separately are highly similar in IQ (0.86), more so than dizygotic (fraternal) twins raised together (0.6) and much more than adopted siblings (~0.0).

Most of the IQ studies described above were conducted in developed countries, such as the United States, Japan, and Western Europe. However, a few studies have been conducted in Moscow, East Germany, and India, and those studies produce similar results. Any such investigation is limited to describing the genetic and environmental variation found within the populations studied. This is a caveat of any heritability study.

Mental retardation

Mild to severe mental retardation is a symptom of several hundred single-gene disorders and many chromosomal abnormalities, including small deletions. Based on twin studies, moderate to severe mental retardation does not appear to be familial (run in families), but mild mental retardation does. That is, the relatives of the moderate to severely mentally retarded have normal IQs, whereas the families of the mildly mentally retarded have low IQ.

IQ score ranges (from DSM-IV)

1. mild mental retardation: IQ 50-55 to 70; children require mild support; formally called "Educable Mentally Retarded"
2. moderate retardation: IQ 35-40 to 50-55; children require moderate supervision and assistance; formally called "Trainable Mentally Retarded"
3. severe mental retardation: IQ 20-25 to 35-40; can be taught basic life skills and simple tasks with supervision
4. profound mental retardation: IQ below 20-25; usually caused by a neurological condition; require constant care

The CDC listed the prevalence of mental retardation across race and gender in a 1991 study: overall, the rate of mental retardation among blacks was 16.6%, compared to 6.8% for whites. Notable is that blacks were almost three times as likely to be labeled mildly mentally retarded as compared to whites. Mild mental retardation is almost never diagnosed until a person enters elementary school, which critics claim lends strong support to the notion that IQ tests are racially biased.

IQ, education, and income

Tambs et al (1989) 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 US siblings, Rowe et al (1997) report that the inequality in education and income was predominantly due to genes, with shared environmental factors playing a subordinate role.

Regression

The heritability of IQ determines the extent to which the IQ of children will be similiar to the IQ of parents. Because the heritability of IQ is less than 100%, the IQ of children tends to "regress" towards the mean IQ of the population. That is, high IQ parents tend to have children who are less bright than their parents, whereas low IQ parents tend to have children who are brighter than their parents. The effect can be quantified by the equation where:

is the predicted average IQ of Mom and Dad's children
is the mean IQ of the population that Mom and Dad come from
h²   is the heritability of IQ
Thus, if the heritability of IQ is 50%, a couple with an average IQ of 120 will have children that average around an IQ of 110.

Brain size and IQ

Modern studies using MRI imaging have shown that brain size correlates with IQ by a factor of roughly .35 to .40. In 1991, Willerman et al. used data from 40 White American university students and reported a correlation coefficient of .35. Other studies done on samples of Caucasians show similar results, with Andreasen et al (1993) determining a correlation of .38, while Raz et al (1993) obtained a figure of .43 and Wickett et al (1994) obtained a figure of .40. The correlation between brain size and IQ seems to hold for comparisons between and within families (Gignac et al. 2003; Jensen 1994; Jensen & Johnson 1994). However, one study found no within family correlation (Schoenemann et al. 2000). A study on twins (Thompson et al., 2001) 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 (Posthuma et al 2002).

The Flynn effect

Worldwide, IQ scores appear to be slowly rising, a trend known as the Flynn effect, so that tests need repeated renormalization.

Sex and IQ

Most IQ tests are designed so that the average IQs of males and females are equal. However, men tend to score higher in the parts of the test that cover spatial and quantitative abilities, and women generally score higher in the verbal sections. Some research has shown that the variance in men's IQ scores is greater than the variance among women's, as seen in other cognitive test scores. This would mean that men are more likely than women to have both very high and very low IQs.

Religiousness and IQ

Some studies claim a correlation between higher IQ, SAT scores, GPA and degree of religious belief. Such claims merit skepticism - especially due to the psychological, sociological and cultural motivations that often prompt such research. While most of the research indicates a negative correlation between IQ and religiousness, this is open to debate.

Health and IQ

Persons with a higher IQ have generally lower adult morbidity and mortality. This may be because they better avoid injury and take better care of their own health. It also decreases the risk of Post-Traumatic Stress Disorder, severe depression, and schizophrenia. On the other hand, it increases the risk of Obsessive Compulsive Disorder [2].

Research in Scotland has shown that a 15-point lower IQ meant people had a fifth less chance of seeing their 76th birthday, while those with a 30-point disadvantage were 37% less likely than those with a higher IQ to live that long.

Practical importance

Research shows that intelligence plays an important role in many valued life outcomes. In addition to academic success, intelligence correlates 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 intelligence and health, longevity, and functional literacy. Correlations between g and life outcomes are pervasive, though IQ and happiness do not correlate. IQ and g correlate highly with school performance and job performance, less so with occupational prestige, moderately with income, and only to a small degree with law-abidingness.

General intelligence (in the literature typically called "cognitive ability") is the best predictor of job performance by the standard measure, validity. 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). The validity of cognitive ability for job performance tends to increase with job complexity and varies across different studies, ranging from 0.2 for unskilled jobs to 0.8 for the most complex jobs.

A large meta-analysis (Hunter and Hunter, 1984) 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 tryout (0.44), experience (0.18), interview (0.14), age (-0.01), education (0.10), and biographical inventory (0.37).

Because higher test validity allows more accurate prediction of job performance, companies have a strong incentive to use cognitive ability tests to select and promote employees. IQ thus has great practical importance in economic terms. The utility of using a one measure over another is proportional to the difference in their validities, all else equal. This is one economic reason why companies use job interviews (validity 0.14) rather than randomly selecting employees (validity 0.0).

Legal barriers, most prominently the 1971 United States Supreme Court decision Griggs v. Duke Power Co., have prevented American employers from directly using cognitive ability tests to select employees, despite the tests' high validity. Using cognitive ability scores in selection adversely affects some minority groups, because different groups have different mean scores on tests of cognitive ability.

Some researchers have echoed the popular claim that "in economic terms it appears that the IQ score measures something with decreasing marginal value. It is important to have enough of it, but having lots and lots does not buy you that much." (Detterman and Daniel, 1989)

However, some studies suggest IQ continues to confer large benefits even at very high levels. 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 (Coward and Sackett, 1990). In an analysis of hundreds of siblings, it was found that IQ has a substantial effect on income independently of family background (Murray, 1998).

Other studies question the real-world importance of whatever is measured with IQ tests, especially for differences in accumulated wealth and general economic inequality in a nation. IQ correlates highly with school performance but this correlations seems to decrease the closer one gets to real-world outcomes, like job performance, and still lower for income. It explains less than one sixth of the income variance . Even for school grades, other factors explain most the variance. Regarding economic inequality, one study found that if we could magically give everyone identical IQs, we would still see 90 to 95 percent of the inequality we see today. Another recent study (2002) found that wealth, race and schooling are important to the inheritance of economic status, but IQ is not a major contributor and the genetic transmission of IQ is even less important. 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.

Economic development and IQ

A controversial book IQ and the Wealth of Nations, claims to show that the wealth of a nation correlates closely to its IQ score. This claim has been both disputed and supported in peer-reviewed papers.

Validity and g-loading of specific tests

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 predicts outcomes such as job performance, social pathologies, or academic achievement. Different IQ tests differ in their validity for various outcomes.

Tests also differ in their g-loading, which is the degree to which the test score reflects general mental ability rather than a specific skill or "group factor" such as verbal ability, spatial visualization, or mathematical reasoning). g-loading and validity are related in the sense that most IQ tests derive their validity mostly or entirely from the degree to which they measure g (Jensen 1998).

Social construct?

Some maintain that IQ is a social construct invented by the privileged classes, used to maintain their privilege. Others maintain that intelligence, measured by IQ or g, reflects a real ability, is a useful tool in performing life tasks and has a biological reality.

The social-construct and real-ability interpretations for IQ differences can be distinguished because they make opposite predictions about what would happen if people were given equal opportunities. The social explanation predicts that equal treatment will eliminate differences, while the real-ability explanation predicts that equal treatment will accentuate differences. Evidence for both outcomes exists. Achievement gaps persist in socioeconomically advantaged, integrated, liberal, suburban school districts in the United States (see Noguera, 2001). Test-score gaps tend to be larger at higher socioeconomic levels (Gottfredson, 2003). Some studies have reported a narrowing of score gaps over time.

While public discourse on IQ testing is generally inflammatory, IQ tests are used ubiquitously in research and education. In general, there is a disparity between the public perception of IQ testing and the opinion of intelligence researchers.

The reduction of intelligence to a single score seems extreme and wrong to many people. Opponents argue that it is much more useful to know a person's strengths and weaknesses than to know their IQ score. Such opponents often cite the example of two people with the same overall IQ score but very different ability profiles. As measured by IQ tests, most people have highly balanced ability profiles, with differences in subscores being greater among the more intelligent.

IQ scores are not intended to gauge a person's worth, and in many situations, IQ may have little relevance.

The Mismeasure of Man

Many scientists disagree with the practice of psychometrics in general. In The Mismeasure of Man, Professor Stephen Jay Gould strongly disputes the basis of psychometrics as a form of scientific racism, objecting that it is:

The view of the American Psychological Association

In response to the controversy surrounding The Bell Curve, the American Psychological Association's Board of Scientific Affairs established a special task force to publish an investigative report on the research presented in the book. The full text of the report is available at a third-party website.

The findings of the task force state that IQ scores do have high predictive validity for individual (but not necessarily population) 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 (again, not necessarily population) differences in intelligence are substantially influenced by genetics.

They state there is little evidence to show that childhood diet influences intelligence except in cases of severe malnutrition. They agree that there are no significant differences between the IQ scores of males and females. 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. While they admit there is no empirical evidence supporting it, the APA task force suggests that explanations based on social status and cultural differences may be possible. Regarding genetic causes, noted that there is not much direct evidence on this point, but what little there is fails to support the genetic hypothesis.

The report was published in 1995 and thus does not include a decade of recent research.