Wednesday 4 March 2015

Girls and boys in Sudan

At a time when I seemed a credible advisor on the topic, aspiring media psychologists used to ask me how to get on television. I replied that only two things were required: having an office near TV studios (my office at the Middlesex Hospital Medical School was in the middle of media land in central London) and having a popular subject, say the sexual fetishes of trauma victims who had been bitten by Royal dogs.

This blog is not a televisual event, but I hope to engage you with the popular and contentious subject of sex differences in intelligence. In part I am doing so because I intended to answer Nick Mackingtosh’s point that Richard Lynn’s work on sex differences had not been generally accepted. Sadly, Nick is no longer alive to participate, and the project was on the back burner anyway, because I thought it required going through many review papers and meta-analytic studies.

Now a paper comes along which gives a good introductory summary of the sex-differences-in-intelligence debate, and adds some new data on a large sample of over 7000 children. By way of background, all informed opinion was that men and women had the same level of intelligence, despite men’s 10% larger brain size. This paradox might have been resolved by showing that those extra neurones were required for the management of the penis. Not so, apparently. The proposal is that the extra capacity is taken up by greater visuo-spatial ability, which begs the question why it does not show up in intelligence measures. Richard Lynn upset the applecart by arguing that boys and girls matured at different rates, and when late maturing boys became men they were about 4-5 IQ points brighter than women. If one accepts that, plus the somewhat larger standard deviation for male intelligence (if one accepts that) then it is easy to see why there are more men at plus two sigma where all the interesting stuff happens. These larger issues to be discussed at a later time.

Salaheldin Farah Attallah Bakhiet, Bint-Wahab Muhammad Haseeb, Inas Fatehi Seddieg, Helen Cheng, and Richard Lynn (2015) Sex differences on Raven's Standard Progressive Matrices among 6 to 18 year olds in Sudan. Intelligence 50 (2015) 10–13.

They say: The Standard Progressive Matrices (SPM) was administered to a sample of 7226 school students aged 6 to 18 years in Sudan. The sample consisted of all the school students in the central sector of the provinces of Al Jazeera Aba, Raback, & Kosti in the White Nile state, approximately 300 km south of the capital, Khartoum. Schooling is compulsory in Sudan from the age of 6 and to 18 years.

The data were analysed for the sex differences on the total scores on the Standard Progressive Matrices and for the scores on the three factors of Gestalt Visualization, Verbal-analytic Reasoning and Visuo-spatial Ability identified by Lynn et al. (2004). To calculate scores on the three factors, a principal components analysis was carried out that showed three factors with eigenvalues greater than unity.

There were no statistically significant sex differences between the total scores of the 6 to 13 year olds, but among 14 to 18 year olds males obtained higher average scores than females and among the 16, 17 and 18 years olds the average male advantage was 0.337d, equivalent to 5 IQ points. An analysis of the data for the sex differences on the three factors of Gestalt Visualization, Verbal-analytic Reasoning and Visuospatial Ability identified by Lynn, Allik and Irwing (2004) showed similar age trends to those for the total scores.

 Table 1 gives the key findings:

image

If I had to talk about it on television I would show the picture below, with IQ on the ordinate. The sex differences are absolutely conclusive at age 17 but indifferent at age 18, and the graph makes that more clear than the table.

image

The picture makes the general pattern clear, with the switchover happening at age 11 when boys start to push ahead a bit, with more gains later. Even without the exceptional 17 year old performance, sex differences are evident.

The authors come to 4 conclusions:

1) no statistically significant sex differences for 6 to 13 year olds, suggesting girls are not disadvantaged in Sudan; (I could posit a later disadvantage)

2) there are sex differences later, broadly but not perfectly in line with Lynn’s 1994 predictions;

3) male advantage is visible at age 12 and statistically significant by age 14 in Sudan, sooner than predicted, and by 16 is at the equivalent of 5 IQ points;

4) the sex differences on the three sub-factors show similar although not identical age trends as those for the total scores. The only major difference is that on Visuo-spatial ability girls performed significantly better than boys at age 11 (.17d).

So, that ends the sex lesson. Lynn’s thesis about male advantage after age 16 was derived from data from European cultures. On any assessment of culture there are significant differences between growing up in Sudan and in Europe, but the same pattern is confirmed in this study. It seems more like sexual dimorphism than cultural constraints, though one can never be sure without comparing many different cultures, and this paper is one piece in the jigsaw.

One thing the paper did not do, because it was not the main topic, was to compare the overall levels of achievement with Western norms. If one takes age 15 as a comparison point then the average score of 28 in the Sudan would be below the 5th percentile on British or US norms. This is roughly in line with the other 12 known studies on intelligence in Sudan.

18 comments:

  1. The reason why this is a muddy topic is that the fundamental problem of getting fully representative samples crops up.

    Dykiert, Gale, and Deary 2008 found that you have more attrition of the lowest IQ males in adult samples. Lynn's and Rushton's findings of a sex difference in mean IQ that appears only in adulthood may be an artifact of this phenomenon. I'd imagine this is an even bigger a problem in a place like Sudan.

    I'm not quite sure how you'd get around this problem given the precision necessary in such a study.

    ReplyDelete
    Replies
    1. Dear JayMan, Agree that population samples are the best, and men "attritate" more than women. Can only suggest we look at the best, most representative samples,

      Delete
    2. Also Flynn (2012, 141-158) find some evidence, that the difference isn't there.

      Delete
  2. "all informed opinion . . . . " Well, received wisdom, anyway. ;-)

    ". . . was that men and women had the same level of intelligence, despite men’s 10% larger brain size."

    Several decades ago, I read somewhere that,while the adult male brain is larger, the number of neurons was about the same for males and females. Further, the neural spacing had something to do with male superiority in visuo-spatial and spatiotemporal abilities. Has this been confirmed or refuted in the past 20 years?

    ReplyDelete
    Replies
    1. Prominent psychometricians like Jensen! Don't know the up to date picture on neurons, but rather doubt it. Will try and check

      Delete
    2. Not sure there's a single paper that calculates the "net" of all the differences but sex differences in neural density in various AREAS of the brain are certainly well reflected in the literature going back well over 20 years.

      See, for example, Witelson et al in Journal of Neuroscience, May 1995, 15(5): 3418-3428.

      As to retrieving the paper I read 20+ years ago, little or no chance, I'm afraid.

      Delete
    3. Thanks for reference. No need to strain for the other one: I accept your summary (if you will accept similar recollection on my part which I cannot trace at the moment, many of which glow in my memory but cannot be found).

      Delete
    4. I would be very careful with interpreting data coming from neuroimaging studies. Small samples + publication bias often give here artifacts. And I would be double careful with data from such studies, reporting sex differences. Only systematic reviews and meta analysis could give reliable picture.

      About cortical density, more recent work, however again - small:

      Alonso-Nanclares et al., 2008

      Delete
    5. Agree care is needed in interpreting neuroimaging studies http://drjamesthompson.blogspot.co.uk/2013/04/not-all-neuroscience-is-rubbish-just-92_18.html

      Delete
  3. P.S. I seem to vaguely recall that the bottlenose dolphin (Tursiops truncatus) was used to support the “neural spacing” thesis. They have “human sized” brains but allegedly, much of the brain is dedicated to echolocation in three dimensional space.

    ReplyDelete
  4. Ah, you dig that up, and I will talk to some brain people.

    ReplyDelete
    Replies
    1. Hey, I'll accept any recollection from you. You're a real scientist. I'm just a retired business executive with an avocational interest in this stuff. I'm here to learn!

      Delete
    2. You are most welcome. I'm learning as well (and forgetting).

      Delete
  5. When I hear Sudan, I think of famine...

    Does severe malnutrition influence the development of male and female brains in the same manner? If food is short, do female children and male children receive the same amount of food?

    ReplyDelete
    Replies
    1. Some evidence girls are left out when food is short. Famines in 1994/1994, 1998, could have been relevant, but mainly affected the South, so authors could look for regional differences to check on this possibility.

      Delete
  6. Girls mature faster on average than boy at least 11-12 years. The testosterone advantage happen after, during adolescence.

    Santoculto

    ReplyDelete
  7. What happens to boys age 17 and why do they fall back at 18?

    ReplyDelete
    Replies
    1. Not sure that they really do, more likely it is simply a variation in the samples, and might not show up in other samples. Meta-analysis would iron out such anomalies, very likely.

      Delete