An Intelligence Quotient (IQ) test is a method that has been suggested to measure intelligence (Chamorro-Premuzic, 2007). This compromises the use of standardized tests that assess one’s verbal, mathematical and spatial skills in determining their score of cognitive ability. This score is converted into an IQ score which is the mental age one has divided by the actual age multiplied by 100 (Maltby et al., 2007). The score is then compared to the standard norm of IQ score distribution where the mean is 100. If a score that is achieved is what one should get for their age, then the mean of 100 would be achieved (Loy, 1998). If one has a mental age above or below their actual age, they would be above or below this average. There has been much debate whether this method of testing intelligence can be a ‘boon’ and provide advantages or a ‘bane’ and has inaccuracies (Block & Dworkin, 1976). Although both sides of the argument will be explained, the aim of this essay is to argue that IQ testing has been useful. The issues that will be addressed include the validity, reliability and ability of the IQ tests that are able to measure intelligence according to an operational definition along with empirical evidence.
The Against Argument
Point One: Lacking Validity
One of the main issues with IQ tests is that they lack validity and predictive validity is one form of this (Maltby et al., 2007). Validity is the measure taken to determine whether the test devised measures what it is expected to. IQ tests have poor predictive validity which is whether these tests can predict future behaviour in a real life setting. Lee & Siegel (2001) report that there is no relationship between IQ and achievement in reading as typically only 20 to 25% of reading scores can be predicted accurately. This shows that IQ tests do not measure a potential of learning and thus assist in predicting future performance.
Construct Validity à cultural bias
IQ tests also have poor construct validity which is whether the structure of the test is a reflection and measurement of the theory of intelligence (Cronbach & Meehl, 1955). Snyderman & Rothman (1987) suggest that there is cultural bias in the questions utilised in such tests (as cited in Mayer, 2007). Mishra (1982) conducted a study to test whether cultural bias was present in three verbal tests in the Wechsler Intelligence Scale for Children-Revised when taken by Anglo and Native Americans. It was found that 15 of the 79 items were biased against the Native American sample. Mishra (1983) also reported that 10 out of 79 items of the verbal subtests of the scale were biased against Mexican-American participants though matched across characteristics like their sex, socioeconomic status and were similar on items that test Arithmetic and digit span.
Construct Validity à Sex differences in testing
Sex bias in IQ testing further proves that the measurement of intelligence has poor construct validity (Selkow, 1984). The scores calculated from IQ tests could be a result of some items that are biased against either one of the genders. Nolan et al. (1977) studied the items of the Kaufman Assessment Battery for Children in differentiating gifted and nongifted subjects on characteristics of race and gender. For gender, 2 items were biased against each of the gifted males and females. 10 items were biased against nongifted Black males and 6 items biased against nongifted females. The experiment indicates that perhaps the sex differences we conclude from IQ tests maybe inaccurate as a result of such biases.
Note: no systematic bias against race or gender was found. It appears that the K-ABC is a relatively nonbiased test suitable for the evaluation of both gifted and nongifted children regardless of race or gender.
Construct Validity à Flynn Effect
Additionally, Siegel (1992) tested whether IQ tests could predict
Siegel (1992) compared two groups of chil dren who had low reading scores. One group, the disabled readers, had reading scores that were significantly lower that those that were predicted by their IQ scores, and the other group, the poor readers, also had low reading scores but these were not significantly lower than would be predicted by their IQ scores. On a variety of reading, spelling, and phonological tasks, there were no significant differences be tween these two groups in reading comprehen sion. These results have been replicated in a study of adults with reading disabilities (Siegel, 1998). It seems that there is no need to use IQ scores to predict the differences between the in dividuals traditionally called learning disabled and those who have equally poor achievement and lower IQ scores. IQ scores do not appear to predict who is able to benefit from remediation (Arnold, Smeltzer, & Barneby, 1981; Kershner, 1990; Lytton, 1967; Van der Wissel & Zegers, 1985; Vellutino et al., 1996). One study (Yule, 1973) found that reading-disabled children with lower IQ scores made more gains than
reading-disabled children with higher scores. We agree with Lyon (1995), who stated, The assumption that a discrepancy between achievement and aptitude( typicallya ssessed using intelligencet ests) isa clear diagnosticm arker for learningd isabilities or can be consid ered a pathognomonics ign isa t best premature, and atw orst invalid. (Lyon, 1995, p. 12) To this point we have argued that IQ tests are inappropriate for identifying disabilities, but our discussion has focused on their use with native English speakers. We turn now to the case of in dividuals whose first or primary language is not English.
- Flynn Effect: changes in norms, the average is increasingwhich includes predictive, ecological and concurrent validity and they also lack reliability in measuring intelligence.
- Reliability is whether the same result would be achieved with the same test if repeated on two different occasions and there is empirical support to suggest that this is not the case with IQ testing (Chamorro-Premuzic, 2007).
Why is IQ testing inaccurate? Evidence for your point?
Point One – Lack of Validity/Reliability: Not measuring what it is supposed to/difficult to generalise
- No predictive validity and ecological validity – lack of relationship between IQ scores and later achievements and later learning
- o Counter-argument: seems to have face validity – ‘appears’ to though not actually
- Other plausible explanations that could have caused the result: external factors – environment, prior practice [lack of reliability, measure at one point won’t be same as another], not everyone has the same experiences and thus cannot measure responses to such questions regarding experiences for people or have innate intelligence [cannot generalise]
- o Internal reliability and validity of measures is high: high correlations
- Classifying people rather than measuring intelligence as shown by lack of correlation with other tests that measure intelligence [lee evidence]
Point Two – Methodological Flaws/Reliability
- Cultural Biases in Questioning
- Sex differences in IQ testing
- Flynn Effect: changes in norms, the average is increasing
- Point Three – Other types of Intelligence à aren’t they important in determining someone’s intelligence and not strictly cognitive?
Implications of IQ testing
- Loy, Jim (1998). My IQ Test. Retrieved 1st January 2010, from http://www.jimloy.com/jim/iq.htm
- Mishra, S. (2006). Evaluation and Assessment