Gender Differences: What Science Says and Why It’s Mostly Wrong

Men and women don’t come from Mars and Venus, but they are undeniably different. Vive la différence! Without it life would be far less interesting. It’s not just the anatomical differences like chest bumps versus dangly bits; there are also differences in psychology and behavior. My daughters and I are frequently mystified by men behaving in ways that seem irrational, stubborn, willful, incommunicative, or inexplicable.

We can only throw up our hands and say, “It’s a guy thing.” It’s perfectly acceptable and accurate to say a certain behavior, ability, or preference is “a guy thing” or “a girl thing” in our society, but that’s only a starting point. It’s descriptive, not explanatory. And it’s certainly not exclusionary: there are guys who do needlepoint and girls who fix trucks. We must go further and ask why a trait is more of a guy thing in our society. Is the trait universal across all societies? Is it built into the DNA of his Y chromosome, is it determined by his level of testosterone, is it a function of gender differences in brain structure and organization, is it a result of cultural influences? Can it be changed? It turns out that answering those questions is extraordinarily difficult. Science has had a lot to say about sex, gender, and sexuality, but most of it has been wrong.

What is “Human Nature”? Does It Differ for Males and Females?

Donald Brown, in his book Human Universals,[i] has compiled a list of “those features of culture, society, language, behavior, and psyche for which there are no known exceptions.” Several of the items on his list describe gender differences:

  • division of labor by sex
  • females do more direct child care
  • husband older than wife on average
  • male and female seen as having different natures
  • males dominate public/political realm
  • males engage in more coalitional violence
  • males more aggressive
  • males more prone to lethal violence
  • males more prone to theft
  • males, on average, travel greater distances over lifetime
  • mother normally has consort during child-rearing years
  • sex differences in spatial cognition and behavior
  • sex (gender) terminology is fundamentally binary

The complete list is available online.[ii]

Human universals describe the consistent features of human nature, but they do not tell us whether those features are innate and fixed (determined by biology) or malleable (determined by culture). It is quite possible that human nature creates these gender differences by establishing customs and educating children in societal norms. And these universals are descriptive, not prescriptive. Even if they have been true for every society so far, that doesn’t mean that they couldn’t (or shouldn’t) be changed in our society.

Evolutionary thinking leads to ready explanations for how gender differences might have enhanced survival of the species. Sometimes these explanations are plausible, sometimes not. It’s plausible that a division of labor with an innate female preference for childcare would enhance the success of breastfeeding and infant survival. But evolutionary psychologists looked rather foolish when they tried to explain girls’ preference for pink. They speculated that women evolved to have a preference for red because this improved their success in the role of gatherer (detecting ripe fruit by its redder color).

They were ignorant of history. The association of pink with girls is purely cultural and quite recent.[iii] Ladies’ Home Journal article in June 1918 said:

The generally accepted rule is pink for the boys, and blue for the girls. The reason is that pink, being a more decided and stronger color, is more suitable for the boy, while blue, which is more delicate and dainty, is prettier for the girl.

Any explanation is superfluous because it turns out girls don’t really prefer pink. A study of 4-5 month old infants showed no gender difference in color preference.[iv] And more than one study has indicated that adult females tend to prefer blue.

Sexism, Overt and Hidden

Much early research into gender differences was motivated by a desire to justify society’s differential treatment of women. In her book The Mismeasure of Woman: Why Women Are Not the Better Sex, the Inferior Sex, or the Opposite Sex,[v] Carol Tavris argues that standards of normalcy were set by males, that studies of gender differences were contaminated by male bias, and that the real differences lie not in gender, but in power, resources, and life experiences.

Sexism has gone underground. On a conscious level, employers may fully accept that women and men are equal and deserve the same opportunities. But the actions of employers don’t match their expressed beliefs. Women are less likely to be hired for certain jobs, and they are paid less. Most people say they support the principle of equality in childcare and household duties, but the reality doesn’t begin to approach the ideal. Women still spend far more time on those duties than men.

Even very young children demonstrate strong gender prejudices. A military woman on her way home from work was wearing her Army uniform and combat boots when she stopped at a preschool to pick up her child. A little girl verbally assaulted her, insisting girls were not allowed to wear those “boy clothes,” demanding that she take them off, and even stomping on her combat boots while yelling “boy boots.” In another anecdote, a boy wore barrettes to preschool. Another boy told him boys couldn’t wear barrettes. In the ensuing discussion, he pulled his pants down and showed his penis to prove that he was a boy wearing barrettes, therefore boys could wear barrettes. The other child said, “Everyone has a penis; only girls wear barrettes.”

Even self-perceptions are stereotyped. Common cultural stereotypes are that girls are worse at math and Asians are better at math. When Asian American women were reminded of their gender before a math test, their scores dropped; when they were reminded of their race, their scores rose.[vi]


Some people still use the words sex and gender interchangeably, but with the rise of feminism and gender studies, it has become politically correct to use “sex” to refer to the biological makeup of an individual’s reproductive anatomy and “gender” to refer to traits determined by social, cultural and psychological factors. People are further divided into the categories of cisgender (where their self-perception of their gender matches their biological sex) and transgender (where it doesn’t). There is a big problem with all of this terminology. It’s often far from clear whether a given trait is determined by biology or culture. There is a two-way street where biology influences gender and gender influences biology. And real people do not fit neatly into a choice of two pigeonholes.

Determinants of sex, gender, and sexuality

Sex (male/female) and gender (masculine/feminine) are not simple binary categories. The reality is far more complicated.

Chromosomal sex. Typically, males have an X and a Y chromosome and females have two X chromosomes. But there are all sorts of other combinations that defy simple categorization. Due to accidents of chromosomal disjunction during cell division, some individuals are XO (Turner’s syndrome), XXY (Klinefelter’s syndrome), XXX, or XXXY. There are genetic mosaics where different cells have different chromosomes (for instance a mixture of XX and XY cells), transpositions where the male sex determinant genes have transferred from the Y to the X chromosome, and other bizarre combinations. One apparent woman with a uterus had 4 different cell lines, XX, XY, XO, and XXY, in her blood and skin.[vii] This gives a new meaning to Walt Whitman’s “I contain multitudes.”

Hormonal sex during fetal development. In congenital adrenal hyperplasia (CAH), a female fetus (XX) is exposed to high levels of adrenal hormone and is born looking like a boy. In androgen insensitivity syndrome (AIS), a male fetus is unresponsive to androgens and is born looking like a girl. In 5-alpha reductase deficiency (5-ARD), androgen levels are normal but an enzyme necessary for male genital development is missing; these individuals appear to be female and may be raised as girls, but at puberty they develop masculine secondary sex characteristics.

Hormonal sex in adults. There may be malfunction of gonads with overor under-production of sex hormones. The individual’s hormones may be naturally produced by the body or may be taken as pharmaceuticals (for instance as a preliminary to sex-change surgery).

Gonadal Sex. There may be testes, ovaries, both, neither, or a hermaphrodite ovotestis.

Sex of external genitalia. There may be a normal or abnormal sized penis, a normal or abnormal sized clitoris, a rudimentary or full-sized vagina, ambiguous anatomy, hypospadias, or other variations. In concealed penis, a penis is normally developed but hidden from view under fat in varying locations. In 1 in 10 million male births, there is aphallia: a failure of the penis to develop in an otherwise normal XY fetus. A person who has had sex change surgery may have a surgically constructed penis or vagina.

Secondary sexual characteristics. These are male or female characteristics that are not directly part of the reproductive system. Breasts, pubic hair in a male or female pattern, facial hair, deep voice, an enlarged Adam’s apple, and many other traits. These can occur naturally or be induced by hormone treatment.

Sex of Rearing.Was the individual raised as a boy or a girl? Even this is far from straightforward. In the case of 5-ARD, if it is known they will develop the appearance of males at puberty, rearing may be ambiguous. Sexologist John Money’s famous sex reassignment patient—a boy whose penis was amputated in a circumcision accident at the age of 8 months—was raised as a boy for 17 months before he was re-assigned to a female gender and raised as a girl.[viii] Sort of. Actually, he could not be raised as a girl normally would be raised from birth. No one actually believed he was a girl; they believed he could be made into a girl for better life adjustment. The experiment didn’t work out so well. He later returned to a male gender role, married a woman, and eventually committed suicide.

Object of sexual desire. In the simplistic binary view, opposites attract. In the real world, individuals of either sex may be attracted to individuals of either sex, both sexes, or neither; to children, animals, fetish objects like shoes or cars; or to sex practices involving suffering or humiliation. The wonderful world of paraphilias[ix] includes such rarities as plushophilia, the attraction to stuffed animals, and formicophilia, sexual gratification from being crawled on by insects. Individuals may either hide these proclivities or act on them.

Gender self-identification. Some individuals feel comfortable with their assigned sex, others experience gender dysphoria, the feeling that they were assigned to the “wrong” sex. These feelings can range from mild discomfort to overwhelming convictions. They can lead to changes in behavior and even sex change surgery.

Social gender. Some individuals conform to the norms of behavior and dress that society assigns to their sex; others adopt the norms of the other sex. They may act like the other sex in every way, or they may just cross-dress, in secret or openly, part time or all the time. (Note: Cross-dressing does not imply self-identification as transgendered nor does it imply homosexual proclivities.) Does the individual want to be called he or she? Does he/she use the men’s or women’s restroom (and do the other patrons recognize him/her as one of them or do they object?)

Legal gender. The laws make different provisions for males and females (draft registration, maternity leave) and may prohibit same-sex marriage. Pension systems in some countries have different retirement ages for men and women.[x] Sex/gender can be legally changed after sex-change surgery. Transgendered persons in transition represent a special case. At what point in the long sex-change process can the sex be assumed to have changed?

A Multidimensional Spectrum

Studies of intersex people showed that gender role or psychological sex could be independent of every physical aspect of sex: chromosomes, gonads, hormones, external genitalia, etc. It gradually became apparent that no innate factor drives sexual behavior and orientation.

Sex, gender, and sexuality are not either/or dichotomies, but a multidimensional spectrum on several axes—biological, social, and psychological. And even the axes are problematic because science has not conclusively shown which traits are biologically determined. Nature and nurture interact and influence each other; it’s difficult to tease out the contributions of each. Each axis has its own continuum, with degrees of strength. A person can fall at the male end of the spectrum on some axes and at the female end of the spectrum on others, and an individual might score high on a scale of feminine interests while also scoring high on a scale of masculine interests. One might have a strong sexual attraction to males and also a strong attraction to females. An individual might have a mixture of XX and XY chromosomes and the anatomical equipment of a female but strongly identify with the male gender and be sexually attracted to females.

It’s fiendishly complicated.

Gender Differences: Now You See Them, Now You Don’t

In Cordelia Fine’s book Delusions of Gender: How Our Minds, Society, and Neurosexism Create Differences,[xi] she writes, “Pick a gender difference, any difference. Now watch very closely as—poof!—it’s gone.”

A study that shows a gender difference is frequently followed by a better study that doesn’t. Repeating a study with slightly modified conditions often completely reverses the findings. In studies showing a gender difference, the difference can sometimes be erased by offering a financial incentive for doing well.[xii]

Women are more empathetic? Gender differences in empathy depend on the awareness of subjects that empathy is being studied. Women may want to be seen as empathetic but may not actually be any more empathetic than men. Self-reports on questionnaires don’t match actual empathetic behaviors, and women’s answers may be “primed” for the stereotype when the questionnaires start by asking their sex.

Babies prefer female faces? Unless they spend most of their time with male caregivers, in which case they prefer male faces.

Men do better on mental rotation tests? But when women were told that “women do better on this test, probably for genetic reasons” their performance on the test equaled that of men.

Men do better on math tests? But when women were told that despite testing on thousands of people, no gender difference had ever been found, they outperformed the men. Cross-cultural studies have shown that the math gender gap, where it occurs, is due to sociocultural factors that differ between countries,[xiii] and that can be changed. The hypothesis that males have more variability in math abilities, hence are more represented at the highest levels, has been debunked.

Both sexes are equally variable.

There have been many studies of girls with congenital adrenal hyperplasia to assess whether their excessive exposure to androgens has made their behavior more “masculine” than that of other girls. Findings have varied, and the whole field is contaminated by a systematic disregard of confounding factors like different upbringing, attitudes of both subjects and influential others due to the sexual ambiguity, the effect of surgeries, repeated doctor examinations of genitalia, etc.

Researchers make assumptions about what is masculine or feminine. In some studies, having fantasies about marriage was coded as feminine but a “positive orientation toward marriage and parenthood” was coded as masculine. Behaviors such as initiating sex, expressing intense physical desire, and masturbation were characterized as masculine. There was no allowance for a conception of autonomous feminine sexuality. Today’s idea that men have a stronger sex drive than women is the exact reverse of the idea in Renaissance Europe that women were sexually insatiable and that men’s rationality enabled them to better control their base desires. Today a man who is extremely concerned with his appearance and dress is thought likely to be gay; in early modern Europe, such a man was thought likely to be a womanizer.

It’s hard to define and measure aggression in humans. Some studies have shown that girls are as aggressive as boys or are aggressive in different ways.[xiv] It seems plausible from animal models that increased levels of testosterone would tend to produce more aggressive behavior. Just think of stallions versus geldings. Castration clearly pacifies stallions; but only if done early enough. If a stallion is castrated after he has learned to behave in typical stallion fashion, that behavior will persist after the surgery.

The whole concept of “sex hormones” is really a misnomer. Androgens and estrogens are also involved in a multitude of non-sex-related physiological processes. Both males and females produce both androgens and estrogens. Higher levels of estrogens are found in stallions than in human females; even when estrogen levels rise during pregnancy, the levels don’t approach that of stallions. Some aspects of male-typical behavior are triggered by testosterone only after it has been converted to estrogen in the brain. Studies have attempted to correlate testosterone with aggressive behavior in men and boys; about half the studies found a relationship, but half didn’t.

There is gender difference in toy preference from an early age, but it may be culturally determined. A six month old girl looks at a pink doll longer than at a blue truck, but that could be because she has been exposed to more pink and more dolls than trucks. At the age of one year, boys and girls offered a gift were equally likely to choose a ball, a doll, or a car.[xv]

Studies of toy preference have been poorly designed and are contaminated by the researchers’ assumptions. In some studies a ball is categorized as a boy’s toy, in others a ball is classified as gender neutral. One study had to stop using Lincoln logs as a boy’s toy because girls spent more time playing with them than with dolls. In a primate study, vervet monkeys were offered a cooking pot as a “girl’s toy.” It’s hard to understand the rationale for that, since neither male nor female vervets cook. Students observing 22-month old children categorized the same behaviors as “male” when they thought the child was male, and as “female” when they thought it was female.

Brain Differences

In 1991 neuroscientist Simon LeVay reported that he had found a difference in brain structure between gay and heterosexual men, with the findings in women matching the findings in gay men.[xvi] Rebecca M. Jordan-Young questioned whether there could be such a thing as a “gay brain.” She wondered how gayness could take a single identifiable form in the brain when it takes such varied forms in people’s lives. In her book Brainstorm: The Flaws in the Science of Sex Differences,[xvii] she questions everything we “know” about male and female brains, gay and straight brains. She found that an unquestioned assumption had guided most of the research on brains and sex: the idea that early hormones create permanent masculine or feminine patterns of desire, personality, temperament, and cognition, that hormones organize the development of brain structures in the fetus and activate behavioral predispositions later in life. She analyzed over 300 studies, and the closer she looked, the less it made sense.

Firm conclusions were based on small and unreplicated studies. Studies found conflicting findings. The choice of what to study and how to measure it reflected researcher bias. Classifications of what behaviors were “male” or “female” changed over time as societal norms changed, and the researchers didn’t even seem to notice.

Male and female brains are different, but are they different in ways that matter? Male brains are larger, therefore men must be smarter, right? Wrong! When scientists realized that brain size did not correlate well with intelligence, they were forced to consider that brain size might simply be a function of body size. They tried hard to find a relative, rather than absolute, measurement that would support their belief in male superiority, looking at ratios of brain weight to factors like height, muscular mass, and even the size of the femur. They failed.

Male brains are 8-10% larger than female brains. It may be that apparent sex differences in brain structure are really just differences between larger and smaller brains. Larger may mean different in other ways: increased size may influence other factors like connectivity. A smaller brain with greater density of neurons, difference in organization, or greater connectivity might be more advantageous than a larger brain. As with penises, bigger does not mean better; smaller brains work just as well as larger brains. Einstein’s brain was only average in size.

Gender differences were found in language lateralization in brain MRI studies, but only in studies with small sample sizes. Large studies and metaanalyses showed no difference.

Gender differences were found in the size of the corpus callosum and the percentage of gray matter, but the direction of the difference was inconsistent. Regional differences in brain structures sometimes disappear entirely after correction for brain volume.[xviii]

The study of sex differences in brain anatomy is still in its infancy. Imaging studies have been enthusiastically embraced, but as yet imaging is a blunt instrument that assumes that an indirect measurement of something like blood flow or glucose metabolism corresponds to actual activation of specific brain processes.[xix] Spurious findings are common, especially when analyzing complex data with a lot of noise in the system: one notorious fMRI scan appeared to show that a dead salmon was “thinking.”[xx] Differences are only averages, we often don’t really understand what we are looking at, and we have no way of correlating the size of a brain structure with its psychological function.

More important, we have no way of knowing whether any differences we observe are innate or induced. Cordelia Fine asks, “But really, when you think about it, where else but in the brain would we see the effects of socialization or experience?”

The whole concept of “hard-wired” differences is flawed. However they start out, brains are malleable and change over time. Due to neuroplasticity, social concepts of gender become an integral part of our brains. Hormones act on the brain, and our hormone levels respond to our actions and our thoughts. Genes are fixed, but their expression is not; they switch on and off in response to our thinking, our behavior, other genes, and our environment.

Outrageous claims have been made for differences in brain structure and what they mean. The prize goes to an educator who told audiences that girls see the details while boys see the big picture because the “crockus” is four times larger in girls.[xxi] The crockus does not exist.

At this point, I think we can safely say there are gender differences in the brain, but we can’t really specify yet exactly what they are.

Gender Neutral Parenting Doesn’t Work

Can we change the influence of society on gender? No matter how gender-neutral parents try to be, they are undermined by their unconscious stereotypes, and children pick up subtle hints from them. Young children are very adept at noticing body language that differs from parents’ words or conscious intent. When parents tell boys it’s all right to play with dolls, most boys still sense that their parents would be happier if they didn’t. As children socialize with other adults and children, they internalize the attitudes prevalent in the society; and peer pressure to assume stereotypical gender roles is very influential. Young children become “gender detectives” at an early age, learning to identify the gender of others and to divide behaviors and dress into male and female categories.

Gender Imbalance in Careers

Why is there a gender imbalance in certain careers, and what (if anything) should be done to balance it? It was once argued that women were underrepresented in fields like science, math, and engineering because they were less capable. When that argument was shot down by the data, it was argued that women simply preferred other fields. But that preference is clearly driven by stereotypes, role models, and implicit associations created by our culture.

Changing stereotypes will require long-term cultural change, which is already well under way. When I graduated from medical school in 1970, only 7% of the doctors in the U.S. were women; today 50% of medical students are women. Can we breathe a sigh of relief and assume equality has been attained? Does that mean stereotypes have been abolished and women in medicine no longer face obstacles compared to men? No, women are still under-represented in surgery and in the higher echelons of the profession.

Fifty percent may not be nearly enough. 70% of the doctors in Norway are now women (and Norwegian women doctors are still complaining about unequal treatment!). If there were truly no cultural impediments, what percentage of doctors would be women? 70%? 85%? What percentage of nuclear physicists would be women? 40%? 50%? 70%? What percentage of nurses would be men? We can’t presume to know, and it would be foolish to speculate. Without better knowledge, it would be irrational to shoot for a 50/50 split in every sphere of human endeavor or use affirmative action to achieve any hypothetically ideal ratio. Instead, we can be aware that inequities exist, we can try to identify and minimize remaining impediments, and we can try to change cultural attitudes. We can work to overcome differences due to nurture and try to level the playing field for everyone. We can even compensate to some extent for differences due to biology (accommodations for pregnancy and breastfeeding). After all that, we will just have to let the chips fall where they may and see how many men and women end up in a given occupation.

As more and more women go into male-dominated careers, society will no longer be able think of those careers as more appropriate for men. Examples of women in those careers will readily come to mind, female mentors and role models will become more available, and stereotypes will gradually change. It will take time. But we can celebrate the substantial progress that has already been made and is continuing to occur.

Gender differences are group differences. They tell us nothing about the individual. They can be ignored when it comes to individual occupational choices. On average, men are stronger than women; but some women are stronger than some men. Military women are now allowed in combat roles. Some women are qualified and want to go into combat; some men are not qualified and don’t want to. Jobs should be allocated by personal preference and ability, not by gender averages or stereotypes.


The science of gender differences is a blooming, buzzing confusion. Any toddler can tell a boy from a girl, but science can’t. There’s no one simple reliable test to determine whether an individual is male or female. We know there are gender differences, but we haven’t been able to pin down exactly what they are. Hormones undoubtedly influence behavior, but we haven’t been able to tease out exactly how, and the reverse is true: behavior also influences hormone levels. There are gender differences in aptitudes and preferences, but they are largely culturally determined. And in the final analysis, they really don’t matter very much for most purposes. Gender differences are averages. The average is irrelevant to the individual and it doesn’t justify job discrimination, stereotypes, or discriminatory political actions.

What are we to do? Reject simplistic dichotomies? Try to describe an individual as 53% male, 43% female and 4% “other”? Try to ascertain their position in a multi-dimensional matrix? Reject the very ideas of sex and gender and stop trying to classify people? Of course not! The binary classification is sufficient for most practical purposes and is very useful.

Real sex differences exist that are very important in medicine. There are clear sex differences in the incidence of many psychological and medical disorders. Boys are more likely to be autistic, to be dyslexic, to have ADHD, and to have Tourette’s syndrome. Women are more likely to suffer from depression. More women have rheumatoid arthritis and multiple sclerosis. We know that men and women have different responses to some medications. In medicine, the knowledge that a patient is male or female is critically important for guiding diagnosis and treatment. And in everyday life, a simplistic classification provides handy answers to practical questions like whether to address a stranger as Mr. or Ms.

We can’t stop categorizing, but we can keep in mind that sex and gender categories are not absolute. Science is not simple. We are driven to categorize, but Nature is endlessly inventive and is indifferent to our taxonomies.

We can’t conclude that there are no meaningful differences in the initial predispositions of male and female infants. We can’t conclude that there are such differences either. The only way to test those ideas would be to study children in isolation from all cultural influences, which is impossible.

In their quest for equal treatment, feminists have sometimes denied that there are any innate differences between men and women. And scientists have over-interpreted questionable and conflicting data to argue that there are many established biological differences. In the scientific search for truth, we can hypothesize that some innate differences probably exist and we can continue to search for them (hopefully using better methodology than heretofore). We must not be afraid to examine differences due to sex, gender, or race out of political correctness or out of fear that our discoveries might be appropriated and misused for nefarious purposes by a group with a political agenda. It is just as foolish to deny the possibility of innate differences as to attribute everything to them.

As Jordan-Young concluded after her exhaustive analysis of the research on gender differences: “Gender is just not a very useful way to group people’s predispositions.”

Note: parts of this article were adapted from two articles that appeared previously on the Science-Based Medicine blog.[xxii]


[i]Brown, Donald. Human Universals. New York: McGraw Hill, 1991


[iii] pink-wasnt-always-considered-afeminine-color-and-blue-wasntalways-masculine/


[v]Tavris, Carol. The Mismeasure of Woman. New York: Simon and Schuster, 1992





[x] Retirement_age

[xi] Retirement_age

[xii]Fine, Cordelia. Delusions of Gender. New York, W.W. Norton, 2010.

[xiii] files/faculty/21_Stereotype_Jan_ 16%2520(PDF).pdf

[xiv] releases/2011/12/11121215 3123.htm

[xv] 2009/06/09/gender-difference-in-math-ability-variability-driven-by-socialinequality-study /

[xvi] 20080916/boys-girls-equal-at-social-aggression


[xviii]LeVay, Simon. Gay, Straight, and the Reason Why. New York: Oxford University Press, 2010.

[xix]Jordan-Young, Rebecca. Brainstorm: The Flaws in the Science of Sex Differences. Cambridge: Harvard University Press 2010.




This article was originally published in Skeptic magazine.

Dr. Hall is a contributing editor to both Skeptic magazine and the Skeptical Inquirer. She is a weekly contributor to the Science-Based Medicine blog and is one of its editors. She has also contributed to Quackwatch and to a number of other respected journals and publications. She is the author of Women Aren’t Supposed to Fly: The Memoirs of a Female Flight Surgeon and co-author of the textbook, Consumer Health: A Guide to Intelligent Decisions.