What Extent Do Gender Differences In Intelligence Exist Psychology Essay

Share this:FacebookFacebook logoTwitterTwitter logo RedditReddit logoLinkedInLinkedIn logoWhatsAppWhatsApp logo 

Sex differences in intelligence have been a popular argument for generations. “Francis Galton had little doubt that men were more intelligent than women” (Mackintosh, 2000, p. 182), whereas Cyril Burt and Louis Terman agree that this should be as straightforward as any concept, and can be answered through empirical evidence. In general intelligence, the findings of various studies have arrived at similar conclusions of either no difference, or men having a slightly higher IQ than women. In specific intelligence, men excel at spatial ability whereas women perform better at verbal tasks. (Lynn, 2005) The major inconsistencies of findings throughout studies have led to the critical analysis of intelligence tests. Implications of these findings will centre on employment and education opportunities, as well as criticisms of methods used leading to suggestions of how intelligence should be measured in the future.

According to Maltby, Day and Macaskill (2007, p. 258), Sir Francis Galton “is the forefather of intelligence tests.” He made the first attempt to study sex differences in intelligence directly. He hypothesised that sex differences do exist between males and females.

Alfred Binet created the first intelligence test for children called the Binet-Simon scale, which could determine the child’s mental age.

Charles Spearman aimed to test an individual’s general intelligence. He was interested in relationships between subtests and found that if an individual scored highly in one subtest, they were likely to score highly in others. He found that these subtests correlated positively with each other. Through these findings Spearman developed his two factor theory. Specific intelligence refers to each type of intelligence that is used for a specific kind of task, such as verbal or spatial intelligence. General intelligence refers to the intelligence that is required to perform on all types of intelligence tests.

In 1938, John Carlyle Raven published his ‘Raven’s Progressive Matrices’, which were free from cultural influences and did not depend of language capabilities

In 1939, David Weschler developed the first intelligence test based on Spearman’s two factor theory called the Wechsler-Bellevue test. This test concentrated on specific abilities and how they correlated with each other to form an overall measure of general intelligence. In 1955 Weschler introduced the Weschler Adult Intelligence Scale (WAIS). These scales included a set of subtests to analyse specific intelligence.

Lewis Terman (1916) studied almost 100 school children between the ages of four and sixteen using the Stanford-Binet test and found that the girls had slightly higher IQ’s.

Charles Spearman (1927) found no sex difference in intelligence, as did Raymond Cattell. Raymond Cattell theorised the differences between fluid and crystallised intelligence, finding no difference on the two dimensions. Fluid intelligence is free of culture and is innate whereas crystallised intelligence is based on cultural experiences and acquired through learning.

J.H Court (1983) conducted a meta-analysis of 120 studies on information provided on the Raven Progressive Matrices. He found mixed results with half of the studies finding no difference and the other half finding a slightly higher IQ in males. However, Court concluded that there were no differences.

Jensen (1998) and Mackintosh (1998) found similar results with the only differences being very small with males having a slight advantage of one to two IQ points.

Anderson (2004) reviewed literature on the Ravens Progressive Matrices and the Weschler intelligence test and concluded that they show no difference.

Richard Lynn and Paul Irwing (2005) did a meta-analysis of 57 studies from 30 countries, covering 80,000 people who gave information on the Raven’s Progressive Matrices. They found no sex differences among children up to the age of fifteen, which does not parallel with Terman’s findings. After the age of fifteen, males gradually scored higher IQ points than women, totalling up to five points when they reached adulthood. However, when looking at the effect sizes of Lynn and Irwings findings, the difference in IQ scores are not very significant. “The effect size allow[s] us to determine the importance of the findings.” (Maltby, 2007, p. 352) An effect size of .2 or below is considered as insignificant whereas an effect size of .8 represents a larger difference. Lynn and Irwing’s finding of males having a slightly higher IQ than females in between the ages of fifteen to nineteen had an effect size of .16, whereas males scoring five IQ points higher than women had an effect size of .2-.3, therefore this finding is not significant.

Lynn and Irwing’s meta-analysis also showed that male’s scores had a larger variance than women. “Irwing and Lynn report that there are twice as many men with IQ scores of 125 and 155, there were 5.5 men for every woman.” (Maltby, 2007, p. 353)

However, Terman (1916) and Herrnstein and Murray (1994) both disagree with these findings. They claimed that they found no difference in the variations of IQ scores.

Although other studies that used the Weschler test agreed with Lynn and Irwing, showing that men had a variance of five percent larger than females.

Even though it has been concluded that there is no significant difference between men and women in general intelligence, specific intelligence is another matter.

The agreed hypothesis is that men and women differ on specific aspects of intelligence. Eleanor Maccoby and Carol Nagy Jacklin (1974) concluded that men perform better on tests of spatial ability and women, tests on verbal ability. M. C. Linn and A. C. Petersen (1985) found an effect size of .44 of males performing better on spatial perception. J. S. Hyde and M. C. Linn (1988) also found an effect size for -.33 of women performing better on speech production, which supports Maccoby and Jacklin’s conclusion. However, due to the small effect size of verbal abilities, Hyde and Linn believe that this is not significant enough to claim that the difference exists.

When considering the validity and reliability of types of measurements, intelligence tests have been criticised. Maltby et al. (2007, p. 298) claims that “intelligence is probably much more than what can be measured by intelligence tests; rather, [it is] the result of the individual engaging in a variety of skills and information within their cultural context.” It would be a challenge to measure a concept such as this which is so dependent upon if the individual is truly willing to take part. Another difficulty is the dilemma of knowing the most reliable measurements. Measures of intelligence produce inconsistencies. Nybourg (2005) states that in half of various studies, there have been no difference found, whereas in the other half, males have a slightly higher IQ, averaging at 3.8. Jensen (1998) found that when certain subtests were eliminated from a study that favoured either sexes, the findings were very different, therefore the findings depend very much on which subtests are used.

In conclusion to knowing the extent of sex differences in intelligence, due to Lynn and Irwing’s meta analysis; there is an insignificant amount of difference in general intelligence between males and females. However, when considering specific intelligence, many studies have shown that men clearly have a better spatial ability, whereas women perform better on verbal tasks. When including each of the specific abilities, they cancel each other out to reach a sum of no difference. Another fact to acknowledge is that male IQ has a much wider spread than females. Females have more of an average grouping of IQ’s whereas males reach wider ends of the spectrum, either reaching a score of 120 and being labelled as a genius, or scoring lower than 50 and being labelled as mentally challenged.

The fact that males have such a large variance in their IQ scores has implications on how IQ should be measured. In the future, IQ should be measured by specific intelligence rather than general due to this large variance belonging to males.

When looking at the implications for males and females in everyday life, these findings will have a big impact on education and employment opportunities.

Leatta Hough (1992) found that intelligence effects various aspects in the work place such as competence and creativity. These aspects will affect future employment for the individual because of how their previous employers will reference them for their prospective employers. The implication of the previous findings regarding males have a slightly higher IQ score than women will directly influence the employer’s decision about who to hire. Employers are more likely to hire men because of their higher intelligence scores.

These findings may also have an impact on what type of job males and females can work in. Due to the findings of specific intelligence, males are more likely to work in jobs that require spatial performance such as construction work, whereas females are more likely to work as a presenter or interviewer which requires high-quality verbal performance.

However, this finding would not have an impact on school children due to Lynn and Irwing’s findings of there being no difference between children up to the age of fifteen. This finding does not however generalise to further education such as applying for universities. Educators and professors of a specific university will want the most intelligent students to study at their university; therefore they are more likely to choose males over females after considering their A Level grades as well as taking into consideration the above findings. However, certain universities, such as Oxford and Cambridge, who interview every candidate maybe more in favour of females due to them being statistically better at verbal tasks. The interviewers knowing this statistic maybe biased before the candidate enters the room which could affect their chances of entry.

In general, the above findings will have an impact on general stereotypes in society. Women are commonly seen as below men in many areas of life, such as the work place. The findings of men having slightly higher IQ’s than women will add to this typecast of women being secondary. Although Lynn and Irwing claim that these findings are insignificant, in today’s society, the fact that men do have, on average a higher IQ score in general intelligence will outweigh the psychologists findings.

The inconsistency of the above findings shows that researching specific intelligence is more accurate than studying general intelligence. General intelligence is a very wide and open research topic, and the precision of studying this form of intelligence is not as clear cut as studying specific aspects of intelligence. Institutions should therefore use specific IQ scores to make decisions, because the specific scores would show the employer or educator where lies the individuals strengths and weaknesses, rather than ordering them in order of intelligence.

In conclusion, when considering general intelligence, males have a slightly higher IQ score than females. However, the effect size for this finding is .2 meaning that the difference in too small to be significant. Specific intelligence tests show that males perform better at spatial tasks whereas women perform better at verbal tasks, therefore cancelling each other out. Due to males and females performing better at different things, their scores average out to be very similar for general intelligence.

This difference does not occur until the individual reaches the age of fifteen. In early adolescents Lynn and Irwing discovered there are no sex differences at all.

Although males score higher on IQ tests, they have a much larger variance in their scores. When looking along the spectrum of intelligence, males tend to score on either end of the scale, whereas females score closer to the average IQ of 100.

The implications of these findings will have a great impact on further education for students, depending on how the university or college base their decisions on. For adults, the findings will impact on what type of job they are more likely to work in.

Due to criticisms of past studies, implications for further studies are that more tests should be conducted for researching specific intelligence rather than general, due to the wide scope of what general intelligence tests can include.


Share this:FacebookFacebook logoTwitterTwitter logoRedditReddit logoLinkedInLinkedIn logoWhatsAppWhatsApp logo 


Approximately 250 words