There is nothing like sex differences in intelligence to put you on the wrong side of half the population. The story so far is that the standard academic opinion on sex differences in intelligence is that there aren’t any, or that they are small, or that the few that exist counterbalance each other. Women are a bit more verbal, men are a bit more mechanical and spatial, but these differences are nothing to write home about, not that men are likely to write home to anybody, tending to leave that pastime to women.
There is agreement among these same researchers that despite equality in means there is a sex difference in standard deviations: women are somewhat more closely clustered around the mean, men more scattered to the world’s imagined corners, with all the variety of an untidy room. There are more extremely foolish and extremely bright men than extremely foolish and extremely bright women.
Like all agreed positions, there are some discrepant findings. For example, it may be a sampling issue, but the usual sex difference in standard deviations does not show up in Romania.
Furthermore, as you may have read in my last post “Women’s brains”, when a large sample of people have their brains scanned, men are 3.75 IQ points brighter than the women, but there is no difference between the two on the standard deviations of intelligence, so that goes against the general pattern of the findings.
Richard Lynn (1994) argued that some of this confusion arises because so many tests of intelligence are carried out on school age children, and since girls mature faster than boys, so they lead in intelligence initially, but when boys finally mature at roughly 15 year of age, men end up a little brighter than women, by about 4 IQ points. This finding has been supported by various studies, though some find male advantage sooner in child development.
Now a new study has been published which shows a male advantage appearing by the age of 10 in Nigeria.
Testing Lynn’s Theory of Sex Differences in Intelligence in a Large Sample of Nigerian School-Aged Children and Adolescents (N >11,000) using Raven’s Standard Progressive Matrices Plus
Yoon-Mi Hur, Mokpo National University, Jeonnam, South Korea
Jan te Nijenhuis, University of Amsterdam, The Netherlands,
Hoe-UK Jeong, Mokpo National University, Jeonnam, South Korea
MANKIND QUARTERLY 2017 57:3 428-437
Corresponding author: Yoon-Mi Hur, PhD, firstname.lastname@example.org
The authors say:
However, Liu and Lynn (2011) observed a consistent male advantage in the Full Scale IQ scores at ages as young as 5 to 6 years in a Chinese sample, and on spatial ability tests of the Wechsler Preschool and Primary Scale of Intelligence (WPPSI) among children aged four and five years in China, Japan, and US. The magnitudes of sex differences were inconsistent as well. While Lynn and Irwing’s meta-analysis demonstrated an average sex difference of 0.33d, two large-scale studies (Lynn & Kanazawa, 2011; Rojahn & Naglieri, 2006) converged to indicate that although sex differences followed the developmental pattern as Lynn (1994, 1999) suggested, the differences after puberty were less than 0.12d and thus concluded that the differences were practically insignificant.
While sex differences in cognitive abilities have been extensively studied in Europeans, Americans, and Asians, there are only a few reports on sex differences in cognitive abilities among Africans. Lynn (2002) administered the Raven’s SPM to 3,979 15- to 16-year-olds in secondary schools in South Africa and found that males obtained a higher mean equivalent to 0.16d among 15-year olds and to 0.31d among 16-year-olds, suggesting that the sex difference increases with age. However, these differences were not consistent across ethnic groups in the study sample. More recently, Bakhiet et al. (2015) analyzed scores of the SPM in 7226 students aged from 6 to 18 years in Sudan. Females tended to perform slightly better than males on the total score up to age 11 years, with a highest d of -.12. From the age of 12 years onwards, however, a male advantage began to appear even though the magnitudes of sex differences were generally moderate ranging from d = .10 to d = .20 with an exception of d = .66 for 17-yearolds.
The present study consisted of 11,164 students (mean age = 13.5, SD = 2.6 years) drawn from three separate samples in the Nigerian twin-and-sibling studies. This sample is far larger than all previous intelligence test results put together, indeed, almost 7 times as large as the best studies previously available. More attention was paid to sampling than ever before. Currently, it is the definitive study of school age Nigerian intelligence.
The children were drawn from public schools, and education officials were involved in picking representative ones, and this was done in Lagos and in Abuja Capital Territory. It may have left out both elite schools and the unschooled, so the exclusions are probably counter-balanced. Intelligence was measured using Raven’s matrices, the best validated and widely used non-verbal intelligence test.
The SPM+ consists of 60 matrix items divided into five sets (A, B, C, D, & E) constructed to become progressively more difficult moving from set A to E. Validity and reliabilities of the SPM+ have been well established (Raven, 2008). As the SPM+ is a non-verbal test, it has commonly been used to assess sex differences in cognitive abilities in diverse populations with different languages and cultural backgrounds.
By the way, if you look at item difficulties on SPM overall, children of all genetic backgrounds have difficulties with the same items (with only 3 items out of 60 showing slight deviations, in two cases Africans having somewhat more difficulty and in the last case Europeans having more difficulty). This means that different genetic groups can have almost identical error patterns while having very different total scores. It is a power difference, not an operating system difference. We can look at those arguments later, in another post.
Anyway, here are the scores for boys and girls in Nigeria.
The pattern of sex differences is a little hard to see, so here it is in a simple graph:
As you can now see clearly, (SPM+ scores on the Y axis, age on the X axis) male advantage is evident by age 10 and increases with age. So, this is another finding which strengthens Lynn (1994) and in this sample puts the age of male advantage back to 10 years of age. This might suggest that Africans mature faster than Europeans, for which there is some evidence, but it seems to be part of a bigger picture of early male advantage in general intelligence. Measured at age 18-19 when students are entering the workforce, or higher education, this is a massive 7 IQ point male advantage. If one takes a broader view, and takes the almost 4000 strong sample of 15 to 19 year olds, the difference is still a 5 IQ point male advantage.
Turning to standard deviations, the average for the whole sample is males 9.2 females 8.6. Clearly, women are slightly more normal. Looking only at the 15 to 19 year olds, the average male standard deviation is 9, the average female standard deviation 8.6 The general pattern of greater male variability is sustained in this sample, though the difference is small.
The authors issue a caution about their findings:
More studies need to be carried out so that a meta-analysis can be performed (Schmidt, 1992), which will allow us to make strong conclusions and examine whether sex differences in cognitive abilities are moderated by ethnic or racial groups. The World Economic Forum (2011) determined that in Nigeria, gender gaps in education, economic empowerment and political participation remain. Moreover, cultural and religious influences foster the maintenance of a ‘son preference’ within the country, which may also influence teachers’ attitudes and behaviors towards boys versus girls. These influences should also be taken into account in future studies.
What does all this mean for Nigerian men and women? If access to professional jobs is based solely on mental ability, then setting a SPM+ cutoff point at 39 (roughly two standard deviations above the mean) you will find that 80% of those top jobs will go to men. So, a fair distribution of employment will seem to favour men by 4 to 1.
Finally, though not discussed by these authors, what do these results mean for Nigeria?
The Standard Progressive Matrices have been the most widely used test in cross cultural studies. Despite an ever-present worry that any test may be culturally biased, close examination on the contrary reveals that the test items show the same characteristics despite the genetics of the test takers.It is as good a test of intelligence as it is possible to apply to different cultures.
Do the results from this massive sample change our view of Nigerian intelligence? Previous studies, based on far smaller samples, with raise the usual doubts about representativeness, had suggested a Nigerian national IQ of about 69, with one sample up to 75. This definite sample comes up with a result of IQ 70 and I think should be considered the best estimate of Nigerian IQ.
Is this result credible, given Nigerian accomplishments? I think we must always question that, and look for discrepant achievements, but I think we probably do not have to revise this estimate, though we will discuss that later, in another post.