Biology, Genetics, recent, Science / Tech, Sex, Social Science

Sex Differences in Cognition

In a previous post I examined the biological and social influences on sex and gender identity. Evidence suggests that biology plays a powerful role in the determination of sex as well as of gender identity, although social forces are also important particularly as they relate to gender role expression. In this essay I’ll examine the evidence surrounding a related controversial topic: whether or not there are cognitive differences between the sexes and, if so, whether they are biological or social in origin.

In what follows, I’ll focus on individuals whose gender identity matches their biological sex. This leaves out nonbinary and transsexual persons, about whom there is far less research evidence. Nevertheless, given that transsexuals tend to have hypothalamuses that match their identified gender not their biological sex, it would be interesting to know if this produces cognitive differences as well. Some evidence suggests that the administration of sex hormones to those undergoing transition does influence cognition in expected ways. Other studies suggest that cognitive differences exist prior to hormone treatment, and that the cognition of transsexuals resembles that of their identified gender more than that of their biological sex (a finding that appears to lend further support to the hypothesis that gender dysphoria is produced by women’s brains in men’s bodies and vice versa).

This essay offers an exploration of mean group differences. Nothing here should be taken to imply that either sex should be excluded from certain cognitive tasks or professions. Nor should mean group differences, which are often quite small, be used to infer the capabilities of any given individual.

Are there Sex Differences in Overall IQ?

Most of the available evidence suggests there is very little difference in overall IQ between males and females. Controversy arises around the question of whether there is more spread (what statisticians call standard deviation or the degree to which most people deviate from the average or mean) among males than females. For IQ tests, the mean is typically set at 100, with a standard deviation of 15. This means that most individual scores fall between 85 and 115. Not everyone, even those with “normal” IQs, is going to score exactly 100. The Gender Variability Hypothesis suggests that boys tend to have more spread or higher standard deviation than do girls—in other words there are proportionally more geniuses and cognitively impaired individuals among males, whereas females cluster closer to the mean. This hypothesis can be very provocative. It was at the center of a 2005 controversy involving then-president of Harvard University Lawrence Summers, who suggested that it might partly explain why males outnumber females in high-echelon STEM (Science, Technology, Engineering, and Math) careers. An outcry ensued, and Summers resigned a year later.

Boys and men tend to be overrepresented among individuals with lower IQs including those with intellectual disabilities. Few people seem to be interested in discussing that. But some studies indicate that boys are also overrepresented in the superior range of IQs (130 and above). In a 2003 study of Scottish youth born in the early twentieth century, girls actually tended to outperform boys until roughly the 115 IQ mark (because there are fewer girls with lower IQs, and more girls in the average to high average range). Only once we pass the 115 IQ mark do we start to see a male advantage. For instance, for IQs 130 and above, boys represent 57.7 percent of the high performers as compared to 50.4 percent of the underlying sample. Biologist Heather Heying recently tweeted a graph from this study. Note that there were still plenty of girls in the highest IQ ranges. So, if IQ differences were the only explanation for, say, STEM Ph.D. discrepancies, we’d still expect 42.3 percent of STEM Ph.D.s to be awarded to women.

A 2016 cross-cultural analysis confirmed the gender variability hypothesis. However, this mainly benefited males in science and math, whereas females scored consistently higher in reading. A 2019 reanalysis of the same dataset confirmed these results, but also suggested that the difference between males and females decreases (but does not vanish) in countries that actively promote egalitarian female participation in the workplace and in education. This suggests that the greater male spread, at least in some abilities, is real and that both biological and social forces play a role.

There are, however, some cautionary notes. One study finds that differential dropout rates in long-term outcome studies may exaggerate the male advantage in IQ variability. Similarly, other analyses suggest that this effect is consistent within the US and UK, but not across all cultures, (although the calculations made by that research group in related contexts have sometimes been criticized). Further, any apparent issues related to sex variability at the top end of intelligence may be due to specific abilities rather than overall intelligence.

Are there Sex Differences in Specific Cognitive Abilities?

It is often said that women and girls perform better at cognitive tasks testing verbal abilities, whereas men and boys have better visuospatial skills involving mental rotation or hand-eye coordination. The evidence appears to bear this out. For instance, in a large sample of Portuguese youth, most cognitive differences between boys and girls were pretty trivial in size. However, mechanical reasoning showed significant differences, with boys outperforming girls. Another study in the US found that girls tend to outperform boys on memory and processing speed tasks, whereas boys once again have an advantage on visuospatial tasks. A very large study of English schoolchildren lends support to female superiority in verbal skills, with males showing an advantage in quantitative abilities.

These sex differences, with female advantages in verbal ability and memory, and male advantages in visuospatial cognition appear to be fairly consistent across samples and across time, subject to the caveats mentioned below. It’s important to note that most of these differences are quite small, and perhaps not worth worrying about. However, males do appear to be at a significant disadvantage in reading and writing whereas comparative male advantages in math and science may explain the greater proportion of males in STEM, at least in part.

Biology, Sociology, or Both?

If there is a sex difference in intelligence variability is it innate or socialized? When we’re discussing human psychology, it’s generally a safe bet to assume that both biological and social forces play a role in shaping behavior. Twelve years ago, a diverse panel of psychologists looked at this issue from every angle—evolutionary, biological, and sociocultural—and found that…it’s complex! Although sex differences in cognitive abilities may not be directly evolved, evolution, brain structure, and culture likely interact to produce various outcomes. The authors conclude that “early experience, biological factors, educational policy, and cultural context affect the number of women and men who pursue advanced study in science and math and that these effects add and interact in complex ways. There are no single or simple answers to the complex questions about sex differences in science and mathematics.”

Some studies suggest that, even in cultures where women are encouraged to participate in highly demanding activities like chess, sex discrepancies in top performers remain. However, other studies suggest that sex disparities favoring men tend to be most pronounced in cultures with overall high levels of sexual stratification, such as women being restricted from entering the workforce. Even in the US, sex disparities have declined over time, suggesting a clear sociocultural impact on such abilities, although stabilization appears to have been achieved more recently, hinting sociocultural forces are only one piece of the puzzle and biology is also important. More opportunities for women in education and occupation tend to at least reduce, though not eliminate, male advantages in some cognitive abilities. My impression is that there has been less emphasis on addressing male deficits in reading and writing, even as males fall behind in attending college.

Training also matters. For instance, some research suggests that women training with fast-paced video games increase their visuo-spatial cognition thereby reducing sex disparities. A study I conducted years ago found few overall male advantages for visuospatial tasks, with each sex better at visuospatial tasks involving items traditionally associated with each sex. By contrast, some beliefs such as “stereotype threat” (the theory that stereotypes such as “girls are bad at math” can negatively influence female performance) are now in trouble, potentially part of psychology’s replication crisis.

That said, some scholars probably go too far in denying the involvement of biology at all. For instance, one paper asserts that mathematics performance “is largely an artifact of changeable sociocultural factors, not immutable, innate biological differences between the sexes.” Although this conclusion is no doubt well-intentioned, it probably goes too far, particularly given how powerfully involved genetics are in intelligence and cognition. Rather, it appears that biological and cultural forces interact in complex ways. Statements that suggest that cognitive differences between men and women are entirely innate are similarly reductive.

Being aware of the science can be difficult in a hypercharged political environment, where hyperbole rules on both sides. For instance, a 2005 Washington Post article made striking claims about brain similarities across the sexes, mainly by pointing to studies of sociocultural influences. This is a bit of a dodge, since few scholars who find evidence for biological differences claim sociocultural influences are unimportant. The author suggests that even sex differences in physical aggression may not be real, pointing to evidence of female equivalence in the perpetration of domestic violence. While that particular data point is accurate, males tend to vastly outnumber females in the perpetration of other violent crimes, a fact that went unmentioned.

In closing, there are several reasonable conclusions we can draw from the current data:

  • There is little evidence for an overall sex difference in IQ.
  • Males may show more variability in IQ, resulting in greater proportional representation at both ends of the IQ scale. However, these proportional differences are probably smaller than is often claimed and don’t fully explain outcome discrepancies, such as in STEM careers.
  • Females appear to be generally superior at verbal and memory tasks, with males superior at visuospatial tasks. Male advantage on visuospatial and quantitative tasks may be one factor in explaining STEM discrepancies, whereas female advantages in verbal abilities may explain females outpacing males in higher education more generally.
  • Sex differences in cognitive ability are most pronounced among cultures with more sex stratification.
  • Genetics have a strong influence on IQ.
  • Sex differences in cognitive abilities are likely due to a complex interaction of evolutionary, biological, and sociocultural forces. Exclusive focus on only one of these is likely to result in an incomplete theoretical model.

A final observation: Much of this debate focuses on perceived differences in ability related to outcomes such as STEM careers. There is also a wide range of literature regarding sex differences in interest, which is also complicated, and which may explain a larger portion of the sex discrepancy in STEM careers. Put simply, many women may have the ability to perform in STEM careers but display more interest on average in alternative high-status careers such as medicine or law.


Christopher J. Ferguson is a professor of psychology at Stetson University in Florida. He is author of Moral Combat: Why the War on Violent Video Games is Wrong and the Renaissance mystery novel Suicide KingsHis forthcoming book How Madness Shaped History will be out in January 2020You can follow him on Twitter @CJFerguson1111


  1. The situation with studies in the field of sexual dimorphism in the West is very similar to those with genetic research in the USSR during the Stalinist dictatorship.
    Vigen Geodakyan proposed his evolutionary theory in 1960–80s (!) but Western scientists remain silent although they risk much less than Nikolai Vavilov, who died of starvation in Gulag.
    I myself began my career as a scientist and left it. There were a lot of reasons for that but now, looking on my past, I know that one of the decisive moments was when I realized that my supervisors are quite capable of lying for the sake of academic intrigue - I was shocked; as a naive young man, I thought the science is about finding the truth.

    For whom it may concern's_evolutionary_theory_of_sex

    PS. Below the copy-paste is placed of my comment to Areomagazine article «The Left is Having an Identity Crisis». It’s addressed to The Left as a whole. The only reason I placed it here is because I feel myself furious by seeing the tragedy in front of my eyes. Sorry in advance!

    A few days ago, my wife and I visited the place where my wife is a regular guest. By the nature of my character, I immediately distinguish people with deviant behavior (I remember, I took a glance at the picture by an artist unknown to me and said that he most likely suffered from schizophrenia – so it was). That’s why I drew attention to a strange young person and quietly asked my wife:
    What kind of creature is this?
    I’ll tell you later – my wife whispered.
    As it turned out, it was a young girl, “positioning herself as a man”. She looked like a frightened puppy, bristling in constant expectation of danger. Small scared puppy.


    We all remember what difficulties we encountered at the age of 12-15 years. We remember what confusion reigned in our minds, how our bodies and our thoughts frightened us. We remember how lonely we often felt. And only after years we began to understand that what was happening to us is happening to everyone.
    The ancients solved these problems by creating rites of initiation. Better or worse but it was helping disoriented teenagers become men and women. In one form or another, these rites have survived to our time.
    And now you, The Left, have come with your genders, queers and hormonal treatment (dirty Mengele followers!). You claim you are who bring “freedom, equality, brotherhood” (I suspect there is no equality you will ever be satisfied).

    But you only bring destruction and death. Ahead of this young girl is only a hopelessly tragic existence. Existence, not life! She will not become a man and she will never be a full-fledged woman as well as thousands of other teenagers will not become healthy and happy women or men. If not for you, they themselves would have overcome their difficulties. But you made it so that these common problems of adolescents forever broke their future.
    Their tragedies are created by you, their ruined lives are the work of your hands. But you despise human lives. In your heads there are only idiotic ideas of an “ideal society” in which there is no place for human being.

    Damn you!

  2. The differences are small but real, even more real is the desire by real men and women (I’m being trans-exclusionary here LOL) to assume more-or-less traditional roles. This is readily apparent to everyone except Marxists and feminists, who are basically the same anyway.

  3. “Nevertheless, given that transsexuals tend to have hypothalamuses that match their identified gender not their biological sex…”

    What is this evidence for this? I did a cursory look at scientific articles and was unable to find anything approaching the level of certainty implied here. Does anyone have any links they know of? Does the author? Thanks.

  4. Expecting 42.3% of STEM PhDs to be awarded to women does not follow from the stated 57.7/42.3 M/F split in IQs above 130.

    To make such a claim (of what would be a proportionate M/F representation of STEM PhDs by IQ) you would have to look at the distribution of IQ amongst STEM PhDs.

  5. I wonder what people will think in the future when they go through the archives to find that in 2019, the absurdly obvious needed to be defended. Hopefully we will not have that long to wait.

  6. Sorry, I’m not familiar with this book, but If you are interested in my opinion, I am quite indifferent.
    Even if racial differences affect IQ, differences in the cultures of communities have a much greater impact. In my opinion, it’s the reason why emigrants from Somalia can be more successful in comparison with African Americans.
    Slavery traumatizes the life of not only an individual person but the entire community, its way of life and its values. In Russia, serfdom existed from the XVI to the XIX century. This determined Russian life in the XX and even in the XXI century; Russia is almost as far from democracy today as she was 100 years ago.
    The major problem is that community life, unlike human life, is measured by generations rather than years and you need constant but careful efforts to fix the problem.

    And let me look at this issue from a different perspective and suggest the following…

    Intelligence is not what is within us; intelligence is what is between us. Neanderthals had bigger brain than we have, I would not be surprised if it turns out that an average Neanderthal was smarter than an average human, but we won, not them. This is the reason why I’m so skeptical about AI.

    PS. I remember, I once said in a dispute about the value of the life of Einstein and the cleaning woman - it makes no sense to compare how much the Everest is higher than K2. They both are 5 miles above Rome’s Capitoline Hill.
    When in the XVI century John Donne wrote “No man is an Iland”, he was understanding this better then we are understanding this now, in the XXI century with all our Facebooks and Twitters.

    Shame on us!

  7. Great point.
    Besides natural distribution of ability, besides interest or lack thereof of the STEM field, there is also one very significant obstacle in obtaining STEM PhDs for women: the process is rigid and antithetical to women wanting to pause the process to raise children. This is one of the reasons I didn’t go into it, among other reasons.

    It takes around 5 years to get your PhD, during which time you are paid near-minimum wages to teach. After that, you generally have to have a post-doc, which is 3-4 years or so, and also underpaid. After that, you land a tenure-track job (which is difficult), and this is where the rubber hits the road–if you don’t get tenure, your career risks being over; at the very least, you drop at least 2 tiers in “status.” During the time you work toward tenure, which takes 6-7 years, you cannot take off for, say, raising children. You will be penalized at hte very least, and at the east will be told you’re not 'taking this seriously enough" and “Maybe this isn’t for you.” One of my ex’s colleagues took a day off for his wife’s birth and was reprimanded for “not taking this seriously enough” by the lead professor.

    Then, if you do get tenure, you are faced with a lifetime of stress and worry in the never-ending pursuit for grants. You have to be a politician in both getting along with colleagues and backstabbing them if necessary; you have to be a boss and run a lab filled with employees; you have to find research that NIH and other funding sources “like,” which very often means you give up your vision simply in order to get the grant.

    The whole process is antithetical to women who want to stay home for a time and raise a family; or for women who dislike hierarchical power games; or for women who just don’t want to spend so much time invested in a career that may or may not pan out while risking their chances for children.

    IQ distribution and interest are factors, but I find way way too many people writing about this subject ignore the very real impediment of the entire *process" of getting a PhD. It’s very significant.

  8. In the authors bullet points, he includes women being both good at STEM skills (whatever they are, presumably math and spatial viualization) and verbal ones as a reason for them perhaps choosing STEM less frequently but ignores the double-whammy of men exhibiting lower verbal skills. As they’re more limited in selection of fields to excel in, it stands to reason they’d disproportionately choose ones where they can not fail.

    Even if 50% of women opted for STEM, more than 50% of men might as well.

    Anyway, these debates are akin to the sex of CEOs at fortune 500 companies, physics professors at Harvard etc: an untoward focus on a tiny slice of the elite population which has little to do with the vast majority of people, male or female.

  9. IQ tests are normalized to the whole population. They’re not normalized to make male and female average scores equal. The averages just work out as about equal. That’s a non-trivial fact, not a tautology.

  10. You got data for your statement? Since there is a sex difference in IQ on verbal and spatial/quantitative reasoning, an IQ test can be made to favor one sex over the other. Therefore, any test is a mix of questions. How can we get exactly at a 100 average for both sexes without calibrating the mix? I.e. if I add one more spatial question to an existing test it will show a higher IQ for boys. You see another way to make this neutral but calibrating?

  11. “the cognition of transsexuals resembles that of their identified gender more than that of their biological sex …”

    This was claimed by transexuals before there were any brain scan studies done. Be very skeptical of such studies now.

    “Most of the available evidence suggests there is very little difference in overall IQ between males and females…”

    When test questions are tried out on test takers before being used for official scores, any question that gives a non-50/50 male/female ratio is thrown out as problematical and never used in testing. So we don’t know if there are differences. If there are, they would have to be very small, but they could be there.

    " The Gender Variability Hypothesis suggests that boys tend to have more spread or higher standard deviation than do girls—in other words there are proportionally more geniuses and cognitively impaired individuals among males, whereas females cluster closer to the mean. This hypothesis can be very provocative …"

    Not a hypothesis. This is an observation that has been replicated many times over decades with different tests in different populations. Let’s not be cute here - if someone is ‘provoked,’ that’s their problem.

    “That said, some scholars probably go too far in denying the involvement of biology at all.”

    They’re called science deniers.

  12. Just a thought: it would be awesome for authors and editors at Quillette to encourage hyperlinking to primary literature on claims like these. I know this is a blog, not a research journal, but considering how difficult it is to get this kind of commentary published in “proper journals” or in the legacy media, it would be helpful for articles like this to provide links to the peer-reviewed literature for people who are interested in digging deeper.

  13. Your whole post was right on as we used to say…The only point I wanted to add is that once a STEM grad gets past the MA/MSc level, they should seriously look into private industry, there seems to be a lot more room for self determination, and much less theatrics. If you are reasonably good in some arcane field, don’t mind jumping companies, have some Intellectual flexibility, and don’t make an unforgivable technical error, you can pretty much write you own ticket.

  14. Too many men, and resources are wasted on fighting over women. You get the instability we see in polygamist societies, where men who don’t have women turn to murder and rape. It’s not evolutionarily stable.

    Too many women, and there become too many children for the men to care for, and resources are stretched too thin for everyone. That is why polygamy only works in times of great abundance, and even then only for the upper echelon. Such conditions do not exist consistently in nature, so this is not evolutionarily stable either.

    Men and women have equal reproductive success because precisely one of each is required for reproduction. Intelligence has nothing to do with it.

    Intrasex competition is not relevant to the question of which sex is more intelligent.

  15. That’s a long way of saying it, but yes, there are differences between men and women. Part of the confusion in the SJW crowd is saying there are no differences between men and women - but if there is no difference, then “transgender” means nothing.

    But I do not think the differences between men and women require differences in public policy, for the most part. And for us as individuals dealing with other individuals, while men and women vary, the variation between any two individuals is greater. Much the same applies to ethnic groups, by the way. Because of the greater variation from one individual to another, we have to take people as they come.

    Sure. I think the overconfidence of men comes from the same root as their greater aggression and risk-taking behaviour: testosterone. This is why the period of highest testosterone, adolescence, is also the period of peak stupidity in behaviour.

    In practical terms, male stupidity is more dangerous than female stupidity, because men’s overconfidence makes them more likely to actually do something stupid, rather than just quietly be stupid. Thus higher rates among men of accidental deaths and injuries, from assault and so on. And men are more likely to bankrupt a company.

Continue the discussion in Quillette Circle

87 more replies


Comments have moved to our forum


  1. Pingback: That's interesting | Tim Worstall

Comments are closed.