Average IQ of students by college major and gender ratio

After all the controversy that arose after I posted my breakdown of college majors by gender last week, I promised myself I’d stay away from controversial gender-related topics for a while. But when I ran across an ETS-curated data set of average student IQs by college major, I couldn’t avoid putting this visualization together. Below, I plotted several college major’s estimated average student IQ over the gender ratio of that major.

The result? A shockingly clear correlation: the more female-dominated a college major is, the lower the average IQ of the students studying in the major. A naive reader may look at this graph and conclude that men are smarter than women, but it is vital to note that, on average, men and women have about the same IQ.


By popular request, here’s an interactive version of the above chart: https://plot.ly/~etpinard/330/us-college-majors-average-iq-of-students-by-gender-ratio/

IQs are typically classified as follows:

  • 130+: Very superior intelligence
  • 120-129: Superior
  • 110-119: Above average
  • 90-109: Average

Considering that many of the female-dominated majors heavily involve interpersonal interactions, my initial thought was that this all made sense: Women are widely known to be more socially-inclined and nurturing than men, so we would expect to see them dominate fields that heavily involve people. But how does that explain the drastic IQ differences between male- and female-dominated fields, if the average man and woman have the same IQ?

The answer comes from the fact that the IQ score here is estimated from the students’ SAT score. This isn’t an altogether unreasonable approach: Several studies have shown a strong correlation between SAT scores and IQ scores. But if we break down the SAT score by Verbal and Quantitative, we see why this IQ estimation is potentially misleading.


If we re-make the first plot against the Verbal SAT score, we see that it’s basically a wash: there’s no correlation between a major’s gender ratio and the average student’s Verbal SAT score.


When we plot the students’ Quantitative SAT score against the major’s gender ratio, we see the negative correlation appear again. This tells us that the original plot is actually showing preference for quantitative majors: The higher the estimated IQ, the more quantitative/analytical the major, and the fewer women enrolling in those majors.

This brings up an interesting question of how valuable the SAT is as a standardized test across all majors, if a higher SAT score is really only indicating that the student is better at solving quantitative/analytical problems. Not all majors require a high analytical aptitude, after all.

Technical bits

Some of my readers requested the R^2 for the above plots. Here they are:

The R^2 on the IQ vs major’s gender ratio graph is 0.601

The R^2 on the Verbal SAT vs. major’s gender ratio graph is 0.019

The R^2 on the Quantitative SAT vs. major’s gender ratio graph is 0.738

The R^2 between Quantitative SAT score and Verbal SAT score is 0.027

For those who want to know what R^2 means: http://en.wikipedia.org/wiki/Coefficient_of_determination

Notice about the IQ data

Since I posted this article, the veracity of the IQ data set has been brought into question. I think StatisticBrain is a fairly reliable data source, but I write this here so readers can come to their own opinion about what this data shows, and how much to trust it.

The data source says “Graduate Record Examination scores” then goes on to list SAT scores. Which is it? According to this comment, the scores listed are pre-2011 GRE scores, which can be found on the ETS web site here. The IQ estimates appear to have been performed separately from ETS, perhaps by StatisticBrain.

So what does this mean for the graphs above? The IQ estimates are representative of students who are in their last year of undergraduate studies (or have already graduated) and are intending to apply to one of the majors. That makes the IQ estimates an imperfect sample, as some students may be changing majors for graduate school. I’d like to see this analysis redone with the SAT scores of students tied to their final undergraduate college major rather than intended graduate school major.

Randy is a PhD candidate in Michigan State University's Computer Science program. As a member of Dr. Chris Adami's research lab, he studies biologically-inspired artificial intelligence and evolutionary processes.

Posted in data visualization Tagged with: , , , ,
  • Nick

    re: But how does that explain the drastic IQ differences between male- and female-dominated fields, if the average man and woman have the same IQ?

    The link to the article sums it up. “males typically outnumber females substantially among high-scoring individuals.”

    It is well known from several studies that while men and women have about the same mean IQ of 100, the variance in IQ for men is much higher than that of women.

    People who attend college are biased toward the upper end of the IQ spectrum (both intuitively and obvious from your dataset); thus we expect men to be “smarter” than women using this subset of data only.

    If you plotted the inverse dataset (occupation by IQ for non-college graduates) you would make an inverse conclusion that women were in fact smarter. I will bet on a waitress against a garbage collector in an IQ test any day.

    • lemmycaution

      The women to men ratio of college attendance is something like 60/40. The men who go to college could easily have higher IQs than the women for that reason alone.

      • dungone

        That could explain why male-dominated majors are filled with high-IQ’d individuals, but it can’t explain why high-IQ’d majors are dominated by men.

        • http://www.twitter.com/klejdys Chris

          Simple; men are more represented at both far tails of the Bell Curve. Men are more likely to be > 130 and < 85.

          • Brutal Honesty

            This has been well documented.

            Most convicted violent felons are about one standard deviation below average IQ and are also mostly men.

            Everyday experience for women has shown them that most of the dumbest people they’ve ever met have been men which is one of the things that makes it so hard for them to understand that most of the smartest people are also men.

          • dungone

            Once again, that explains why male dominated majors have high IQ’s, but it doesn’t explain why those majors are male-dominated. There’s a difference and it’s important.

            There are a handful of female-dominated majors that are also correlated with high IQ’s in this data, such as art history. So why do men flock to CS while women flock to art history? As the first clue, I’d suggest looking at why physics majors often end up becoming quants in finance. The reason is money.

        • muhprivilege

          It’s all about the money. Why do people do art history? Why do they teach English? Why do they do social work? Not for money.

      • bobol

        thats just not true, in the U.S there are actually an even amount

    • Deborah A. Dixon

      In response to your comment about IQ and college, there are a couple of big grains of salt that this neat line of best fit needs to be taken with:

      Firstly, economic opportunity plays a big role in why that line looks like it does. All other things being (un)equal, men with IQs in the range of many of the ‘low IQ’ majors displayed here could do a much better job securing their economic futures in the skilled trades, and do not *need* to attend college, or attend college for a year or two and drop out to apprentice in a wide variety of fields, some of which are quasi-engineering such as autoCAD drafting, etc.

      This drives the skewed gender ratio among undergraduates in favor of females (and as a college professor in the sciences in South Texas I witnessed firsthand what happened to our gender ratio when the Eagle Ford Shale fields started getting worked down here) – and as a result, the *average* IQ (based on a larger slice of the population than of males due to the aforementioned difference in opportunity) among college females drops relative to that of males.

      Ancillary to this is that most of the equivalent IQ level ‘majors’ that are likely to be male dominated that we do not see on this chart have been farmed off to vocational/technical schools and so are missing from the ‘majors vs IQ’ chart – also resulting in skewed looking data erroneously suggesting that college females congregate ‘low IQ’ majors or are ‘of lower IQ’. Dedicated nursing and teaching technical schools used to be common, but now these students are lumped in with more traditional academic majors.

      Additionally, one needs to look at the *proportion* of college students in each of these majors; how many Education vs Mathematics majors are graduated every year? The absolute number of physicists and mathematicians in this country is tiny relative to that of teachers, nurses, and retail sales managers (which is where most of those Psychology majors are going to end up) and people in these fields are at the high end of the IQ spectrum simply because of the demands those fields make on their intelligence.

      Finally, the poor performance on the quantitative portion of the SAT, as well as the ‘math gap’ seen starting in late elementary school among females, is a distinctly American phenomenon – this ‘quantitative gap’ is greatly reduced to nonexistent in Asia and the Scandinavian countries. Even Mexico, despite its culture of machismo, does a *much* better job teaching basic quantitative skills to its students than we do in the US (a phenomenon I am constantly reminded of watching my immigrant students relative to US-educated ones, performing basic scientific calculations, and was particularly evident when I was teaching at an institution on the US/Mexican border).

      And *finally* finally, as a woman of a certain age who was discouraged from taking Calculus in favor of a *typing* class, and then advised not to bother taking calculus based physics (I vigorously fought both of these suggestions and got my course placements, but this was in the early ’80s so not *that* long ago), one should not discount, despite the cheerleading for STEM in this day and age, the constant, subtle and not-so-subtle, discouragement young girls and women are subjected to throughout their training in these careers. Being that engineering and the quantitative sciences *are* hard, just as large numbers of the merely ‘above average’ intelligence men abandon higher education for something that will be more fruitful, women abandon engineering, physics, and mathematics unless they’re hopelessly smitten, because nursing is more financially renumerative, and unlike in engineering, where they may be the only one in a classroom in the US, they’re not effectively locked out of the support resources as study groups and camaraderie that make the hard work bearable. I can spot these gals a mile away because they bust the curve in my Nursing Microbiology classes, and breeze through the dilutions and viable plate count calculations while the bulk of the students leave me wishing I drank after a hard day of working problems over and over ;).

      TL;DR: Differences in economic opportunity drive gender-based differences in the makeup of undergraduate student bodies. The US stinks at teaching quantitative skills from every conceivable angle. Human beings are social animals and congregate where they will receive support. Engineering and science are hard and demand high quantitative IQ. The difference in the tightness of the bell curve of intelligence is a real thing, so where you take the slice matters.

      • http://www.randalolson.com Randy Olson

        Easily one of the best comments here. Thank you for commenting Deborah.

        Regarding taking the number in the students in a major into account: Interestingly, there is only a weak correlation (R^2 ~= 0.06) between the number of students in the major and the average IQ of the major. Although, I do see a qualitative trend of the popular majors (Business, Psychology, Education) being near the “Average” end of the IQ spectrum here (~110). But there are also some fairly “popular” majors in the high IQ spectrum also, for example Engineering, Economics, and the Mathematical Sciences.

      • Ryan

        Great comment.
        This pretty much explained it for me.

        Summarizing your comment

        1. The stats do not reflect the absolute numbers of students in each major. This explains the misguiding correlation.

        2. Women are less likely to choose their majors based on their high IQ. (subconciously)

      • muhprivilege

        What may have discouraged you from the 80s is long gone. Women are encouraged now more than ever to participate in STEM roles. Whether they do or not is their own choice.

        You are wrong about nursing being “financially remunerative.” You insinuate higher education is a necessity among men and those who deviate make is acceptable for women to not pursue a math intensive career choice at all. Learning to a be a nurse no doubt has its challenges but it pales in comparison to becoming a physicist or engineer.

        • Sandy

          “What may have discouraged you from the 80s is long gone.”

          Ha ha no. I’ve tutored female college students in the past few years who got the “girls are bad at math” treatment in their schooling. Women friends who go to software meetups have repeatedly gotten “are you in marketing?” or “are you someone’s girlfriend?” There may be encouragement of girls to go into STEM — I’ve got other friends who run Python workshops for women — but it’s still pushing uphill against a mass of sexism that’s far from ‘gone’.

        • Victoria

          It’s not long gone by any stretch of the imagination. It’s just much, much more subtle–and therefore, harder to identify and terminate.

          The current trend of encouragement is… well, encouraging. But lasting change is slow to arise, and most female-led STEM campaigns are still a drop in the bucket compared to the overall trend of unconscious bias.

          • Class of 88 CHE

            Nothing will keep the women that like STEM out of those fields and anything will keep those capable that don’t really like it in the field. Ladies grow a thicker skin. You are working way too hard to find what probably isn’t there. Even if bias is ever so slightly, subtly there, so what. My advice. Get out of your heads and go live life on your terms. And Jim in the post below – I always out-performed and it never occurred to me to think in the way you outlined. Honestly we fill young women and men full of gender neuroses these days. It’s a wonder men and women can sit in a room together with out a fight or flight stress response over anticipatory manifestation of gender discrimination. I see interesting data but I don’t see a problem or a judgement. Men are better in math. Sounds like a data supported logical inference that doesn’t bother me but I am better than most men at math and science. Even if there is no difference in Quant test scores in Sweden that’s only translating into not quite 9% of current enrollment in Tertiary STEM degrees. So what. Just because you can do something equally well after high school education doesn’t mean you like it and want to do it for 40 years. If you like something and practice it you get better at it. I’ve been working since graduation in the field for a global company been in engineering, R&D, manufacturing, and IT – what a ride. Worry less, live more.

      • Jim

        Great comment! I got mad and wrote a longer angrier reply to someone in another thread, but thought you might be interested (if you haven’t seen it before) to look at the work on stereotype threat. http://www.reducingstereotypethreat.org/

        Basically, the research on stereotype indicates that if you belong to a group that society thinks is bad at something (say women and math), and you care about performing in that domain, you’ll perform worse in that domain b/c you’re anxious and/or trying hard to prove the stereotype wrong. If you remove the salience of the stereotype, or do other types of affirmation interventions the effect goes away. In other words, women do tend to perform worse than men on math tests, but only when the test is presented in a context that activates the stereotype of ‘women are bad at math’. If you give men and women math tests, and tell them before hand that the tests do not show a gender bias, then there is no gender difference in performance.

      • empirc

        The ‘math gap’ at the average is distinctly American, but the gap at the extremes of the bell curve is not. Higher male variability is quite well established. Male-to-female variance ratios are generally thought to be anywhere between 1.10 and 1.21. This seems small, but at the 99th percentile a male-to-female variance ratio would result in a 2:1 ratio of men to women based purely on IQ. The ratio blows up even further past the 99th percentile.

    • AndrewK

      Who cares? The average person in this country has one testicle and one breast. Good luck finding them. This is nothing more than mental masturbation. Don’t you have anything productive to do with all that brain power of yours?

    • AndrewK

      Who cares? The average person in this country has one testicle and one breast. Good luck finding them. This exercise is nothing more than mental masturbation. Don’t you have anything productive to do with all that brain power of yours?

  • Jason Graika

    I think some part of the problem is the bias that is is the SAT exam. Your data set for IQ is just based on SAT scores. The SAT is well known to be biased towards males.


    However the correlation still looks too strong and other factors are likely in play. One guess is that because more females attend college, the average female IQ in college will be lower.

    • Bob

      correlation =/= causation

    • Joe

      “The SAT is well known to be biased towards males.”

      “Well known”

      Yes, one site mentions that guys do 35 points better and 3 points better on the verbal. 3 points!

      Give me a break.

      • Jonathan

        Regardless, proving that men do better at a test doesn’t imply that it’s biased towards men unless you cling to the a priori belief that men and women are equal. Numerous studies have shown they are not and many commenters here have explained why observed variance in men and women is consistent with men testing higher on a college bound test.

    • Mike

      >The SAT is well known to be biased towards males.
      This is an extremely weak point that only survives because it is intended to thwart sexism.

  • http://inside.mines.edu/~pconstan Paul Constantine

    Mathematics is Pareto optimal!

    • nsodju

      No, it could have fewer women.

  • Jeff

    Well, since I don’t believe men are inherently smarter than women either, the reall issue is what IQ really measures. There are tons of discrepancies in IQ scores between various groups, varying largely on cultural differences. A lot f researchers haves looked into it. If I remember it correctly, the upshot of it all is that IQ tests measure how well conditioned a person is to the westernized culture and it’s academics.

    • Jonathan

      Everything measures something. A test that measures how well you will do in the academic environment of your culture? Sounds useful to me! But, I guess what matters most is what you believe.

  • GA

    Interesting. I think there is a relationship with the aspirations to do certain types of work. The majors align nicely with western stereotypes of what men and women are ‘meant’ to do. It may be when people start focusing on their careers that they read less and less stuff that isn’t relevant to them anymore. Psychology majors read psychology papers etc, which makes them less fluent in abstract ideas such as physics and maths.

    Also, IQ tests measure more and more our ability to abstractly represent and manipulate information. Accordingly quite abstract subjects like math and physics get people who score high on IQ tests.

    It may be that women have a tendency towards careers, for social reasons, that are practical and about interpersonal relationships, which may be self-selecting towards majors that are less abstract and thus score lower on IQ tests?

    • http://www.randalolson.com Randy Olson

      It may be that women have a tendency towards careers, for social reasons, that are practical and about interpersonal relationships, which may be self-selecting towards majors that are less abstract and thus score lower on IQ tests?

      After looking at the SAT score data, I’ve come to this same conclusion. I think it just goes to show that IQ scores really don’t fully capture intelligence — only the ability to work and think in abstract.

      • Kelly

        I was always told, and this seemed true when I took the IQ tests for gifted, that the test primarily measures pattern recognition.

        As far as that goes you will see a need for more pattern recognition in STEM majors.

        There is the psychological thing as well girls tend to be more harsh on themselves, so if they are not 100% sure they might not try STEM. This explains the low percentage of women in STEM.

        These IQs are also averages, which is affected by the sample size. STEM majors usually have a smaller population and because of the intimidation factor have a higher concentration of people with high IQ. Other majors tend to have a much larger population with a broader range of people in them.

      • http://www.twitter.com/klejdys Chris

        What this really shows is people who are clamoring for more women in STEM fields should STFU. As the kiddies might say, women are not as interested in STEM fields because human nature.

        Also, IQ tests may not be perfect, but they are damn close. So close that Harvard, Yale & Princeton use them as does the US Military (ASVAB & AFQT). They work.

        • Gaby

          Women are not as interested in STEM fields because human nature? Do you have any evidence for this? It is more likely environmental factors. Women are discouraged from following these fields from early on. I suggest reading this report http://www.aauw.org/resource/why-so-few-women-in-science-technology-engineering-and-mathematics/

          I am a woman in science with an IQ of about 140. There isn’t anything “different” in my nature. I was born in a home of encouraging people and was motivated by them to do well in math and explore my surroundings. I am sure many women (and men) can do well in STEM (and any field) if encouraged.

          • http://www.randalolson.com Randy Olson

            Agreed Gaby. I think it’s crazy to say that women are genetically predisposed to dislike STEM. Perhaps the best evidence for this is the fact that, as Deborah A. Dixon mentions in another comment, this STEM gap is nearly non-existent in several other countries. The STEM gap is much more likely to be caused by cultural norms in the U.S.

          • Jeff


            Holy cow you are dense as lead and should not use your IQ as a point of argument because it only makes you look solipsistic, a terrible female trait. The USA is as free as any nation can hope to be, this enables people to self-select their future to a very high degree. Tuition is cheap and the government will pay your way. If you are super-smart like you, then schools will offer you many scholarships.

            The simple fact is that, compared to men, women, on average, are more interested in people and nurturing and social concerns than they are in object orientation and focus. Nothing is preventing women from swooning over the Toyota TS040, and some do, but with far less frequency than do males. Since interest in the TS040 is a purely personal choice, and one that can be made alone, in a closet far from evil men, the absence of female fawning desires over the TS040 speaks volumes about what women want.

            I love the TS040 and similar objects, but I do not pretend that most men do. I know men who love objects similar to the TS040, and they think a high fraction of other men share their interest. I think these men are idiots because they lack a truth orientation to see that high object orientation is a limited behavior. You are like these men; you think because you are interested in X, that other women are like you. The fact that our society is open, that education is accessible, that interest in object orientation can be seen in personal hobbies and online viewing habits, and that women choose to participate less frequently than men, is telling. But if you insist that you are right, and that evil men are putting women down, then you can bravely enter the world of HBD, which is the big leagues for finding out that you are nothing special and humans are nothing more than genes and those genes code our behavior. Here is a paper on personality differences in men and women: http://www.plosone.org/article/info%3Adoi/10.1371/journal.pone.0029265

            You cannot win because truth is manifest. Women in nations like India go into tech, because it is a way out of the unfortunate situation that is India. However, once in the USA where people have far greater options than most nations, their choices diverge. None of this is to say that you are abnormal for your interests. But you are like my friend who thinks other people like the TS040 when truth is that unlike you, they don’t even know what the TS040 is.

            Here is the predecessor TS030, I am sure you will watch this dozens of time, just like me: http://youtu.be/eRECqhIZwIA

        • May

          It’s not “human nature”. 66% of girls of 4th grade girls say they like science and math. Then there is a sudden drop off in female’s interest in STEM in high school, college, and beyond. Obviously this is not the result of some concerted decision made by all females to suddenly lose interest in a subject which they previously found interesting and full of possibility. It’s a result of being told that STEM is for the guys, and that girls should focus on more “feminine” academic areas; it’s a result of hearing about women in tech who are accused of only “receiving” certain positions or promotions because of their gender, or whose abilities and qualifications are constantly being challenged because they are women.

          No one is forcing STEM careers down girls throats. Groups like NCWIT are simply trying to give girls an equal opportunity to consider a career in a STEM field, and encouraging young women to not eliminate the possibility of a future in technology or engineering simply because of their gender.

          Source on the stats above: https://www.youtube.com/watch?v=XP3cyRRAfX0

          • Jeff

            You are wrong, it is human nature. For when given a choice, women choose differently. Randy above shows the limitation of his thinking when he agrees that because women in less free nations, where capital availability is lower, and economic self-sufficiency is less probable, choose tech that it somehow indicates an equal interest in tech across the sexes. The simple fact is that men and women are very different. Probably no nation on earth gives people a better chance at success than does the USA; the fact that women avoid tech like the plague speaks volumes. But if you disagree, fear not, for genetic research will soon show you to the be the fool common sense reveals you to be. In the meantime, here is a paper on personality differences: http://www.plosone.org/article/info%3Adoi/10.1371/journal.pone.0029265

          • Jim


            Not sure you understand the word ‘solipsism’. For someone who seems to like to claim you’re making well reasoned arguments, you also seem to lean quite heavily on ad hominem attacks quite a bit too. Why would you ever call someone stupid when you’re trying to write persuasively? All you’ll do is make it incredibly unlikely they’ll even begin to consider your arguments.

            Which takes me to your arguments – they’re bad and you should feel bad. Your primary argument seems to be that they are genetic differences between men and women, and that environment plays NO role in individual differences whatsoever. This is a view that no scientist in their right mind would agree with today. Since you like big 5 personality papers, maybe look at this search: http://scholar.google.com/scholar?q=gene+environment+interaction+personality&btnG=&hl=en&as_sdt=0%2C5. Environment plays a HUGE role in personality.

            Actually, the best thing for you to read would be the very interesting work on stereotype threat. Stereotype threat describes situations in which a commonly held societal belief influences a group of individuals performance ONLY because the belief alone. For example, if you give a Math test to a randomly selected group of men and women, you typically find that the men outperform the women. However if you give another group of randomly selected men and women the same test, but just tell them before hand ‘this is a test that men and women typically perform equally well on’, men and women perform EXACTLY the same way. This phenomenon can be seen on ‘academic achievment tests’ for students that our culture typically think have low academic achievement (eg black or hispanic students), or for white males performing tests of ‘athletic performance’. I actually you’ll think this is vey interesting: http://www.u.arizona.edu/~jeffs/golf1.pdf. Just read the abstract/skip to figure one, if you want a quick summary. But remember, this same effect is seen w/ gender & STEM field performance.

            For more info: http://www.reducingstereotypethreat.org/

            And Chris – learn what causality is – this graph doesn’t conclusively show anything except the current distribution of SAT scores by gender. I actually think it’s a really stark reminder that we need to do MUCH better encouraging young women to pursue STEM fields. Also learn how to make an argument without the phrase ‘STFU’. You’re embarrassing yourself. Actually, as a man and as a person you and jeff both embarrass ME.

          • Gaby


            Yes, women and men can be different and their personalities vary as well. And yes, genes control many things. However, the environment controls genes (and importantly in the case of behavior it also controls neural circuits). Complex behaviors have been hard to trace back uniquely to single genes because of this.

            Saying that the fact that women chose differently because of their nature does not make sense. You aren’t evaluating the environment they were exposed to while growing up that lead to this choice. For example, you like cars. Maybe when you were young you played with cars and were encouraged by someone you looked up to. Some girls grow up in homes with strong gender stereotypes and are not allowed to play with cars or simply not exposed to them. When these girls grow up they may show little interest for cars while you still have this interest. This does not mean that they naturally don’t like cars. These choices are in fact purely personal as you say, but they have been influenced by the opinion of others. The same thing can happen with career choices. As May said, these girl’s interest and ability in STEM drops in high school, they are not born with an innate preference and this is why it is probably not in their nature.

            Few things are uniquely genetic or environmental. In personality it is usually an interaction of both and it is hard to dissect these two apart. In the particular case of career choice, if you read the report I posted it seems to be to a great extent the latter. Now if you show me articles where they show that men and women have different abilities and personalities due to GENETIC differences that can not be overcome by environmental influences that lead to career choices I would love to read it.

            And please if you are going to reply, avoid the derogatory adjectives towards us or females. It is rude and I am simply trying to have an enriching discussion. Thank you.

          • Gaby


  • L

    the problem that I have with this is that you reduce gender to an oversimplified binary, presumably based on biological sex. however, even from an academic perspective, gender is not necessarily reducible to biological sex (en.wikipedia.org/wiki/Social_construction_of_gender_difference). gender encompasses many factors that transcend biological binaries. I assume that your data simplified male to “people with male reproductive organs” and female to “people with female reproductive organs” to female. to do so, in my opinion, ignores the fact that gender is at least largely socially constructed and collapses a highly dimensional feature space (i.e. some space of personality traits) down to a binary space without justifying why such an approach is appropriate.

    • lemmycaution

      This study is based on SAT data that asks for your future major and for your gender. Presumably, transgendered students can select their gender according to their gender identity so it is not strictly based on genitalia. It doesn’t matter because there are not very many transgendered students.

    • D

      That’s cute, too bad your biology and your hormones responsible for, I guess “intelligence” aren’t dependent on how womanly you feel or not.

      “gender is at least largely socially constructed”

      No, it’s really not. Keep this shit on tumblr and outside of scientific discourse.

      • ida

        Gender socially constructed. Chances are, you don’t have children.

      • K

        Guessing you assumed this was a woman and started taking out all your male aggression on her for her opinions, yes? Chill out, bro. She’s right that this issue extends beyond biology into the societal realm. The beliefs that society holds about gender does have an effect on how an individual comes to think about their value to society. With regard to schooling, there are studies that prove that a girl’s perception of her worth and ability will effect her test results (no doubt because women process and attend to language differently than men). Stop blowing your load over biology. The innate differences in our brain structure are not in question- it’s the attitude toward these differences that need to be examined. The reality is women are expected to take part in the workforce, which means we need to restructure some areas of our society with respect to gender. Adjusting your language might be a good way to start, lest you’re content with the idea of being surrounded by upset women who have power. Which is what’s happening by the way, and it has nothing to do with biology and everything to do with the reaction to women in the workforce. Your assumption that society has no bearing on gender ability is made false by neuroscience which has proven that women are acutely aware of their surroundings and grossly effected by language. You might not have made the connection (another heightened feature in the female brain), though it should have been obvious to you by now since using language techniques with women is quite popular in the manosphere. Luckily there are actually scientists doing the work, instead of arrogantly trolling and trying to silence everyone who has a difference of opinion.

        • G

          K, I like how you shoot down D for assuming L was a woman (which D didn’t even actually do), you then go on in the very next sentence to assume L is a woman yourself.

          Then you try to downplay the effect (correct use of effect) biology has on gender, instead advocating for how environment affects (correct use of affect) the genders. While at the same time making several references to how women are biologically more inclined to XYZ than men are, belittling your own point in the very motion of making it.

          It’s quite fascinating yet extremely saddening.

      • http://www.twitter.com/klejdys Chris

        I wish I could give “D” 10,000 +1s. Perfectly said.

    • K

      This post certainly begs the question…

    • Bryanb

      Did you just reference Wikipedia in an attempt to disprove a collection of data you disagree with? Bring that SJW BS back to tumblr and stay away from academic discussions. You bring nothing credible to the table.

      • A

        When someone cite wikipedia, it’s customary to actually check if the article itself references good academic material. While the article does have quite a few problems, there is a vast amount of material on how gender is at least partially a social construct. You’re contributing far less to any academic discussion by dismissing L’s argument for a shitty reason than their argument has.

        Nonetheless, as previous commenters have stated, the genders are self reported, and the small amount of trans individuals is unlikely to impact any of the major conclusions. Those are good reasons to argue against L, yours are not.

    • Mike

      >act that gender is at least largely socially constructed

      This is not a fact. It is conjecture fueled by ideological bias.


      • Mike

        http://www.fairtest.org/ is a pretty biased source.

        They state: “Twice as many males as females achieve SAT scores over 700. If the scoring gap were caused solely by the larger pool of females taking the exam, females should still attain the same percentage of high scores as males. In fact, the opposite is true: the gender gap is largest in the highest score ranges.”

        They seem unaware of the fact that greater male variability in IQ scores explains that.

        Further, I believe that the SAT discards questions that show a bias toward one or the other sex — so the sex differences are actually minimized.

        I know that this is the case for IQ tests — they were designed specifically (via a statistical procedure called item analysis) so that males and females have, on average, about the same overall IQ scores.

        On tests of cognitive aptitude that have not been designed to eliminate sex differences in overall scores (Raven Progressive Matrices) one does find greater average sex differences.

        There are areas where women score higher than men, such as verbal fluency and object location memory.

        • K
        • K

          The proof of a difference in brain structure and function has already been explained by the field of neuroscience. This is baffling to me, because usually whenever something is scientifically proven the public changes their attitude. We already have the facts- the public just isn’t aware. It’s absolutely a political issue.

          • Jim

            Neuroscientist here. Jury is still way out on what anatomical gender differences mean. We still don’t understand how gross anatomical structure maps to function in almost any context (e.g. disease contexts), let alone differences in gender. And that’s even before you make the jump about genetics and not developmental/epigenetic/environmental factors being the primary influence on function (or structure for that matter).

  • D

    The reason seems rather obvious: the IQ distribution of men is bimodal, while the IQ distribution of women is unimodal. i.e. You get really smart men, and then see get some REALLY dumb ones. The dumb ones are not reflected since many dont show up in the dataset for college students since they didnt get into college.

    • http://www.randalolson.com Randy Olson

      I had that thought also, but I don’t think there are enough men on the “genius” end of the spectrum to cause such a large bias in the majors like this.

      • Mike

        those averages shown aren’t in the genius level, and if you were in the computer science/astronomy/physics fields, and made it past the first year or so, the data here is very believable, at least in my experience.

  • Robert

    Women don’t have to earn a living. They can go to college to find a man. So why work the brain muscle too hard?

    • Leigh Sheneman

      “Women don’t have to earn a living. They can go to college to find a man. So why work the brain muscle too hard?”

      Woman do in fact have to earn a living. Not only is the cost of living such that two incomes are necessary to live comfortably. Some of us actually like to work and think quite poorly of people of this mindset. I am simply amazed a male has the guts to vocalize such an outdated and bigot point of view in this day and age.

      • Justin

        Altough tone is hard to read in text, I’m fairly sure that the commentor was being satirical. We can’t always take a comment like this at face value.

        • Morgan

          In this age of the men’s rights movement, the poster could be completely serious.

          • Raymond Martin

            So what.

      • Michael

        I make enough on my own to live comfortably because I am an intelligent male and don’t have to give no bitch no dough

    • Chris

      Please don’t breed.

  • jim terwiliger


  • jayleigh

    I have had this discussion with many people about IQ. My wife and I excelled in grade school through high school. She graduated valedictorian. We both went to college. I have a BS in Electrical Engineering and an MBA. She has a BA in Math, and a masters in Computational Engineering (high concept math/computer science mix). Neither of us have ever taken an IQ test. Who takes them? Why? What are the questions? If I never encountered one through academia, what validity does it have. I have always been curious and chalked it up to the test that the scientologists give that feed into people’s ego.

    • skeptic

      If you read the article, the IQ scores are based on SAT scores.

      • Zack

        Actually if you read the article, Skeptic, it says SAT scores are CORRELATED with IQ… not based on them. So much for being a smart ass eh?

        • A

          How’s Skeptic being a smartass? Jayleigh seemed to call into issue WHO was taking these IQ tests, and Skeptic’s response, I think, was meant to clarify that the given IQ scores are based on actual SAT scores (and this is done, because as you note they are correlated).

        • greg

          Don’t eat from the troll. Statistically any thread of comments on the internet will degrade exponentially to gobbly gook.

  • Dean

    If you could make a trellis chart of this data, it may be t(r)elling.

    I think it is plausible that the result is explained by a culture that discourages women from analytical pursuits. I also think that more analytical => more discouragement towards women. As this cultural quirk is overturned, we can expect the most analytical majors (ironically)to be the last to give.

    W/r/t “The higher the estimated IQ, the more quantitative the major (with Philosophy being an odd exception)” if you replace ‘quantitative’ with ‘analytical’ then philosophy is less of an exception..

  • John

    A lot of people are analyzing these data wayyyyyy outside the scope of what it intends. The results show a very narrow conclusion: that among college students, male-dominated majors have women in them with higher IQs.
    No where in the data does it implicate that IQ correlates to sheer intelligence, or that men are smarter than women in these fields because of IQ variance (how could you possibly get that from the data?).

    Watch your scope and don’t make saltatory inferences from data! Please and thank you.

    • John

      Just two clarifications:

      1. I meant male-dominated majors have people* in them with higher IQs.

      2. The data suggest that men do, in fact, have higher IQs than women given the IQ variance. That could mean that, in a physics class of 10 with two women, 5 males could have an IQ below 100 and 3 of about 150 (i.e., large variance) and the two women of about 120-130. This plausible example demonstrates that the “higher IQ” statement should not be set in stone. My main point: don’t jump to conclusions. Just take the data for what they’re worth.

  • herpderp

    Men have greater IQ variance, and the low IQ males are not in Uni.

  • Scott

    The amount of fidgeting to get around the truth is quite humorous. Sure, gender identification extends past genitalia, but I bet that single predictor explains the vast majority of the variation between “male” and “female”, as you would have them described. Someone did make a good point that abstraction competency is also only one part of intelligence. It can be equally argued that the individuals who scored highest on the predictive IQ based on SAT inferences are among the more emotionally/socially incompetent out there. Furthermore, there are biological differences in the brain structure of males and females, resulting in higher emotional intelligence for women than men. Facts, by definition, are not slander.

    As far as standard IQ goes, the most demanding of the sciences are male dominated. No need to wag the dog.

  • https://twitter.com/bahniks bahniks

    An easy explanation is that women are on average smarter than men in most of the fields. That way it would be possible that quantitative SAT corresponds to IQ fairly well while both sexes may still have same distribution of scores. It would be possible to test that if you have data for both sexes separately.
    Simple example: you may have 20% of women in physics with average IQ 155, 80% men with average IQ 130, making the average IQ for physics 135. Then you may have 80% of women in education with average IQ 115 and 20% men with average IQ 90, i.e. average IQ for education 110. This way you may get the pattern of results in the graph.

  • StonerHippie

    If you think philosophy is an “odd exception” in the way you describe, clearly you have at some level never understood what philosophy is all about – pure logic and reasoning. Philosophy involves having a robust understanding of logical principles (which it fathered) — principles which became the mathematical axioms which the fields of math and physics rely on today, so it makes sense they are grouped together. It also makes sense philosophy is slightly lower than mathematics on the chart because it always draws its fair share of lower performing (on average) stoner/hippie students who have found it to be their “goto” major under reasoning such as “I don’t know what to study so I’ll study philosophy because it’s about cool unanswered questions and my friends would think I’m selling out if I became a major other than environmental science or community planning”.

    • Philosopher

      THANK YOU. I was reading that and though “This person doesn’t know what philosophy entails.” Technically he’s right that it isn’t “quantitative, but if you just substitute the word “analytical” it makes perfect sense. Philosophy is the study OF analytics.

    • Kennon Gilson

      See my comment on philosophy above 1/26/15

  • Cindy

    Can you outline the majors that have an average of <100 IQ?

  • Flynn

    Given that IQ is not static over a person’s lifetime, the obvious question is: are these “IQ” figures (and SAT scores, for that matter) for these students taken before their college education, or after? It would hardly surprise me to find that mechanical engineers and computer scientists are better at puzzle solving, like the kind you’d find on an IQ test.

    The URL of the source is “iq-estimates-by-intended-college-major”, which suggests that not only is the IQ score taken from *before* their college education, but there’s no certainty that their intended college major ends up being their actual college major (one of my professors who kept track of such things said that the vast majority of students did not graduate in the field they originally thought they would), or even if they graduated at all.

    That is, it’s essentially a self-reported survey. Everyone knows engineers make more money than social workers. Ask a bunch of incoming freshmen what they think they can do. The ones who have been told they’re smart and good at math pick engineering. That’s hardly surprising, but it’s also not very informative.

    Could it be that boys who had higher math scores wanted to become engineers, and girls who had higher verbal scores thought they wanted to become teachers and social workers, *because* they had not yet been to college and seen what these fields actually entailed?

    I’d be interested to see before-and-after IQ scores for college students, and also before-and-after college majors. As it stands now, these charts are very misleading.

  • XorFish

    I think a huge part of the difference could be explained with the number of males and females that go to college.

    The difference is quite huge. Ranging from 33 to 50% more women than men that go to college.

    It could very well be, that a lot of men that could major in female dominated fields drop out or want to do something what they deem more rewarding.

  • Jennifer Bartholomew

    There are a number of factors that combine to affect career and educational choices, including aptitude (which is where I’d include IQ scores as a subfactor), interest, salary aspirations, and social factors such as status/prestige etc. You’ve written insightfully in the past about the disincentives men face in choosing careers such as nursing and education, given the assumptions people tend to make about men in female-dominant professions. Similar factors affect women’s choices. Your interest and aptitude for a subject both have to be considerably higher than average to make it come out as the best choice when you would be a minority in the field.

    So overall I’d say what this trend implies is a vicious circle: women avoid STEM careers because they know that other women are avoiding those careers, and being the only woman (or one of a handful of women) in a room full of men is often uncomfortable or embarrassing enough to outweigh other factors in favour of a profession. And the same would be true for men who would have otherwise liked to become, say, nurses, or early childhood educators. Social factors are particularly influential on teenagers, which is where we are in life when we’re choosing our college majors. This would mean that as the percentage of women rises within a profession, the number of female undergraduates willing to major in the subject would also rise, so I suspect that a graph of the % of women in historically male professions chronologically from the point where women are first permitted to enter the profession would show that the growth rate is exponential. Or maybe j-curved, if the earliest adopters have very negative experiences which are made public – something we have seen in the media for many professions, including IT and the military.

    • Dave

      Sure, that explains the difference in male-to-female ratio, but it doesn’t explain the strong inverse correlation between SAT and major.

      What we need is to know if the SAT was taken before or after. Because of the rigorous nature of some of these majors, one could argue that the majors themselves will increase the test taking ability of the students. If that’s the case, then the “social construct” argument could be valid because the major selection would affect test scores, not the other way around.

      I highly doubt that the SAT was given after graduation though… I think that it’s exactly what other posters have mentioned: men have a higher variation and therefore make up a much larger percentage of the high SAT scoring spectrum.

      Because highschool classes should be similar for both men and women, I don’t think you can argue that cultural or social factors play into this. A test is a test, and the background and preparations for men and women shouldn’t begin to diverge until after college starts.

      • Jennifer Bartholomew

        The point I was trying to make in the first paragraph is that the variables are interrelated (aptitude, interest, $, social factors) and that where one is weak, another must be significantly above-average to compensate. This study just considers two variables in two of the factors (aptitude, measured here by SAT scores/IQ) and social factors (measured here by just one factor, gender). I am not familiar with SAT scoring so I’ll refer to IQ – apologies if the analogy doesn’t hold up to SAT.

        Assuming Career X and Career Y will both have negative social factors for an individual because of the way his/her ethnicity or gender will affect their experience in the career. If Career X is one in which the average member of the profession has an IQ of 100, then 50% of all people could be considered above-average at that profession, and it will not take long for the critical mass of new entrants to the profession to be achieved. Career Y is one in which the average member of the profession has an IQ of 135, so less than 5% of the population has the potential to be above average. Career Y thus would have an extremely slow rate of growth for any ethnic or gender group in which the social factors of that profession are perceived as negative, and it would take a very long time for the critical mass to be achieved, if ever. So it would be perfectly natural to see more Career Y professions in a low-diversity quadrant of the scale, and more Career X professions in a high-diversity quadrant of the scale.

        Another way to test this is to see if the same pattern holds for various ethnicities as it does for gender. If it does, then the argument “women like socially oriented careers better than math and science” would have to be restated as as “men of Z ethnicity, and women of all ethnicities, like socially oriented careers better than math and science”. I’m not sure as many people would be comfortable with that hypothesis. However if it turns out that careers with high IQ thresholds are also less ethnically-diverse than careers with lower IQ thresholds, then that would support my suggestion.

  • A

    How does the data correlate with graduation rates for each major? If only 25% of engineering majors graduate (I think that’s somewhere in the ballpark), and 70% of communications majors do, wouldn’t we expect the average IQ in the more selective field to be higher?

  • John

    The IQ system of measuring intelligence is inherently biased towards quantitative reasoning and observation based data retention. It is, essentially, a measurement of how close you are to the idealized version of intelligence as dictated by the makers of the test, which is inherently biased against different forms of intelligence such as emotional intelligence, social intelligence, physical intelligence and creative intelligence. If you are trying to discern how smart someone is, a test should encapsulate all aspects of mental intelligence, not just number based reasoning.

  • Dan

    Philosophy isn’t an “odd exception.” Philosophy is hardcore analytical training. Moreover, philosophy is the father of science, which was called “natural philosophy” until just the last ~200 years.

    • Ryan

      Excellent point

    • Kennon Gilson

      See my comment above.

    • Raymond Martin

      Wrong. You are misinterpreting the term “natural philosophy”. Science went on its own path because philosophy was insufficient to explain reality.

  • Pingback: Average IQ of students by college major and gender ratio [OC] | Go Elect Yourself!#$()

  • Graham Ketteridge

    We should clearly end women’s suffrage

  • Matt

    None of this takes into account the actual number of people in each job. The majors that are closer to even, gender wise, skew lower than the trend would suggest. This data seems to show a correlation between IQ and major more so than IQ and gender. Lower IQ females gravitate towards certain majors, but that says nothing about where lower IQ males end up. They may be more spread out as males are encouraged to a wider variety of majors compared to women. Historically women are discouraged from entering Science, Math, Engineering, etc. majors. This results in only the most motivated and intelligent women pursuing these majors which skews that data higher. Nothing contained in this information indicates causality, just a random assortment of data.

  • Sarah Ametxa

    HAHA WOW!! This is wild!

    My current major is Pre-Engineering that’s 6 blasted semesters of MATHEMATICS!! 6 Physics classes, 4 chemistry classes 4 electrical engineering classes to name a few!!

    ” The higher the estimated IQ, the more quantitative the major (with Philosophy being an odd exception), and the less women enrolling in those majors. Many women tend to avoid mathematics due to social stigma, lack of exposure during childhood, etc., and this fact is ultimately reflected in their SAT performance.”

    I do what I can to optimize myself. For the longest time I’ve had the wrong perceptions on math. In the past I dislike math only because I had poor self image in my own abilities to master mathematics and I didn’t have the motivation or support from my peers to enable me to be confident in mastering it.

    Now my perception has changed I like that doing math problems creates new neural pathways and expands cognitive capabilities in the brain.

    I came back to NYC to enroll @QCC to achieve my #1 goal: Improve my Metacognition in Mathematics. I’m doing so by obtaining my associates degree in Pre-Engineering.

    • Erik Ponce Morales

      Who cares…

  • Ronald Roma

    Kudos for including R^2 values. But please please put them on the graphs. Don’t make me match them up later.

    Basically all you’ve said here is that high achievers in math at age 16-17 go on to major in mathematical fields. Big whoop. Everyone already knows that boys do better on the SAT math.

  • Mike

    Men and women have about the same average IQ.

    However, they differ in the amount of within-sex variability in IQ scores. There are many more men than women at both the bottom and the top of the IQ score distributions. The greater proportion of men at the very high levels of IQ can explain the trend.


  • Sweden as a bad example

    @Randy Olsen

    A naive reader may look at this graph and conclude that men are smarter than women, but it is vital to note that, on average, men and women have about the same IQ.

    No, Mr. Olsen, a non-naive reader will conclude that women are inferior, because this result reconfirms what every man sees with his own eyes: Women are inferior and it shows in nearly every test result, whether IQ tests (women score 5 IQ point lower on the average), knowledge tests (women score even worse), income tests (self-employed women earn half as much as self-employed men) or car driving tests (women have 5.7 accidents per million miles, men only 5.1) and so on.

    Nearly the only tests where women are better than men are tests that have to do with babies or cleaning, for example face recognition or sensitivity to smells.

    I guess you added this statement about “naive readers” because you are afraid that in the current climate and the current degree of femastazation of academia it’s wiser to not point out politically incorrect facts. After all Lawrence Summers was fired for completely harmless remarks.

    • Dan

      Stop generalizing and while you’re at it, take a Logic 101 class. To address one of many bad points you make, it is not a fact that “women are inferior” to men. Some women are inferior to some men by some measures. Some men are inferior to some women by some measures.

      You, for example, are certainly among the men in the latter statement.

      • Sweden as a bad example


        >Stop generalizing

        Learn the difference between “Dogs have 4 legs” and “All dogs have 4 legs”.

        • Dan

          > Learn the difference between “Dogs have 4 legs” and “All dogs have 4 legs”.

          This is why you need a logic class. There is no difference between the two. “Dogs have 4 legs,” without any modification, means exactly “All dogs have 4 legs;” it does *not* mean “Most dogs have 4 legs” or “Dogs on average have 4 legs,” despite the way the language is misused colloquially.

          Meanwhile, both equivalent statements—”Dogs have 4 legs” and “All dogs have 4 legs”—are false, in just the same way as your previous generalizations are false.

          • Sweden as a bad example

            >There is no difference between the two.

            Oh, I see, the word “all” has no meaning.


            Was nice talking to you, Dan, which stands for “Danielle”, I suppose.

            • Dan

              > Oh, I see, the word “all” has no meaning.

              Add another straw-man to the list of fallacies you’re racking up.

              Of course “all” has meaning. In this context, it simply adds nothing extra to the proposition “Dogs have 4 legs.”

              Your arguments are “inferior,” to use a word you like.

          • Ryan

            OMG TOLD!

    • A

      Troll harder. Look at the comments for more than a second and you’ll find a good amount of explanations for the trend seen here. A difference of IQ in 3-5 points, which found in meta-analyses of many studies (in some of which women score higher than men) does not indicate any large scale patterns capable of explaining the above graph. Even more so, the higher variability of male IQs has been suggested (and when it’s possible, tested) as an explanation for this observed difference. Fewer men are tested in IQ studies, and these are often self selected – only those who are already in the upper half. Cherry picking random “tests” in which women perform worse is a hilarious tactic; men also have much higher rates of dyslexia, why not throw that in as another inferiority of women?

      I also love how you have decided that facial recognition is clearly something only needed for child care, and sensitivity to smells only for cleaning (how long do you think humans have lived with any form of cleaning and hygiene?). By the same token, I’ll declare that male driving performance is only needed to flee from crimes, and IQ tests are only needed for the mental gymnastics of sticking your head so far up your ass that you can ignore the world around you and pretend that “women are inferior”

      • Sweden as a bad example


        >Look at the comments for more than a second and you’ll find a good amount of explanations for the trend seen here.

        Yes, you can explain it all away. Just like Star Trek fans and bible worshipers explain all Snafus and inconsistencies away.

        That’s what White Knights in science do, too.

        “Oh, it’s not female inferiority, it’s just because X, Y and Z”.

        Yet study after study comes in and it shows a clear pattern: Women are inferior. No amount of White Knighting can help you with that.

        >Cherry picking random “tests” in which women perform worse is a hilarious tactic; men also have much higher rates of dyslexia, why not throw that in as another inferiority of women?

        I know that men are more represented at the lower ends of tests, but that is misleading.
        First of all homeless men do not balance out the billionaires.
        And second: Even the retards are useful and can dig a hole. And who cares what the IQ of George Michael is? Even if George Michael had an IQ of 80 he would still be a good music composer.

        Women on the other hand (even the high IQ females) are nearly always worthless, and if they are not worthless then they have less children, hence are detrimental to the gene pool of future generations.

        >I also love how you have decided that facial recognition is clearly something only needed for child care.

        Which I never stated.

        I stated that the nearly the only tests in which women outperform men are tests that are needed for baby handling or cleaning or similar.

        Women are designed by nature to be less intelligent (too much brain power wastes energy needed for lactation etc) and are designed to stay at home and do monotone tasks like cleaning babies and cooking.

        So to clarify my statement: Women are not inferior as such, just as men aren’t superior as such. Women are SUPERIOR baby washers, lactators, floor cleaners etc. and men are inferior baby handlers, but are superior scientists, artists etc.

        Women’s desire to sleep more, to suffer from scoliosis, to be more cowardly, to be less intelligent, to have weaker bodies, to have periodic belly cramps and so on are actually superiorities because they help women to be better stay-at-home moms.

        • A

          Didn’t try hard enough.

          >That’s what White Knights in science do, too.
          Because, you know, you can make science say whatever your agenda is, and not just give you clear results. Clearly never done any science, huh?

          >Women on the other hand (even the high IQ females) are nearly always worthless, and if they are not worthless then they have less children, hence are detrimental to the gene pool of future generations.
          Clearly no understanding of the gene pool or how distribution of reproductive effort works.

          >too much brain power wastes energy needed for lactation etc
          Lactation, staying at home, etc. are all far less energy intensive than say, going out hunting, or constructing things or even just developing the extensive amount of muscle men do. By your own logic men should be significantly dumber than women.

          Keep sticking that head up further your ass, but I don’t think it will help with your lack of understanding of how logic or science work.

          • Sweden as a bad example


            >Because, you know, you can make science say whatever your agenda is, and not just give you clear results.

            The results are clear: Women perform worse. You are the one who doesn’t want to accept that this is not some funny glitch but is simply the consequence of women being inferior (or rather: consequence of women being the superior stay-at-home sex).

            >too much brain power wastes energy needed for lactation etc Lactation, staying at home, etc. are all far less energy intensive than say, going out hunting, or constructing things or even just developing the extensive amount of muscle men do. >By your own logic men should be significantly
            dumber than women.

            Re-read your nonsense, maybe it will dawn on what a White Knight you are.

          • A


            >Claims women need to save energy for “lactation and etc.” so can’t be smart
            >Men need more energy just for natural development than women do
            >Doesn’t see how by that logic men should be dumb as shit

            Go back to /pol/

          • Sweden as a bad example


            > Go back to /pol/

            Thank you for reminding everyone that there is far more freedom of speech and far more contest of ideas at 4chan than in science.

            By the way that’s partly because of the high influx of women. Every area goes downhill as soon as women infiltrate, whether science, politics or education.

            You know why.

            • A

              People circlejircking is hardly “more freedom of speech and contest of ideas”.

        • Mike Hanson

          *cough* resentful virgin *cough*

  • Terrymac

    I’d have to see the distributions to make this definitive, but the tendency of men to greater variance could explain this pattern. A person with 110 IQ would probably struggle in math or physics. My guess is that a 130 IQ would be needed to feel comfortable in that territory – so how bad is the skew in the 130+ realm?

    As for amelioration of math anxiety – several women have told me that Montessori-based approaches were much better than “traditional” approaches. I could say a lot about math instruction, but it would become tl;dr.

  • James

    Since high-IQ is (by definition) rare, a large number of people in a major should correlate to a lower average IQ. Given an even IQ distribution by gender, but women gravitating to more popular majors, this chart is accurate (but misleading). I think you’re seeing a dual self-selection of a few smart males choosing science majors, while large numbers of males and females from across the spectrum distribute to the other majors.

  • Longhorn

    Why are you correlating choice of major (free will) with SAT scores (intelligence)? And why are you focusing on one gender? This is not only hasty, but it’s faulty, and it begs the question of intent. You’re a computer scientist. You should not be posting correlational studies and inviting the public to form untaught assumptions about gender and education. These issues belong to the field of psychology.

    • A

      I think he is inviting the public to think about the cause of the correlation, not claiming anything beyond that.

      As an aside, “begs the question”? Doesn’t that mean it’s a circular argument? Sorry, not a native english speaker, just confused as to how you would think this is circular.

      • Dan

        “Begs the question” is a commonly misused phrase. It really means “assuming the consequent,” or, as you put it, engaging in circular reasoning. It’s been misappropriated by the masses, however, to mean something like “prompts the question.”

  • Pingback: Infographic: The More Female Dominated A College Major Is, The Lower The Average IQ | Thought Catalog()

  • George Gale

    If, as many recent thinkers have held, mathematics follows from logic, then philosophers are not “an exception.” Rather, their high score follows from the priority of analytical logic over quantitative subjects.

    • Kennon Gilson

      See my comment above. Philosophy is a tough major/minor.

  • Pingback: Only L<3Ve @ ItMakesMeStronger.com()

  • anonymous

    Can I request a plot that shows the IQs of all comp sci PhD candidates and a highlighted dot for where you fit amongst this? along with an anova test to show the significance of the trend and 3 peer reviews (one in statistics, two from comp sci) to confirm the approach and findings?

    • http://www.randalolson.com Randy Olson

      Sure, you can request it.

      • http://GonzoEcon.com Tony Lima

        Where’s the “like” button for this reply?

  • Jason Zimba

    I think you should make this a bubble chart, because the number of physics/astronomy majors is minuscule in comparison to other majors.

    The most natural explanation for the top chart would be that smart women tend to choose more randomly among majors, while smart men tend choose from among a few.

  • andy

    While avoiding stereotypes and gender bias based on iq, id hazard a guess that with the ratio showing lower iqs show a direct correlation to sat scores, id say toward the lower end of the profession chart, those jobs are more emotion based than logic based, and id like you to argue in a general sense women tend to think emotionally over logically? Im not saying some women dont think logically and some men dont think emotionally, but maybe this plots something thats intangible.

  • Jim Clark

    We need separate plots for males and females to interpret this, as well as numbers for each major. Smaller numbers for certain majors would itself introduce different average IQs if aptitude is part of the selection.

    Interestingly it appears that high female majors at the far right include a lot of education students. Perhaps would vary across cultures.


  • anonymous

    Randal, here’s recent (2011-2013) data on GRE scores by intended graduate major from the ETS website itself:
    The table starts on page 30. Hope this helps.

  • Pingback: Bag of Randomness |()

  • NotANumber

    Perhaps, I should wear a tag with my SAT scores, GPA, and major.
    Sorta “Gattacaa”-lite version of the movie. Saves time from
    the three questions or dog butt sniffing for dominance at parties
    and bars:
    1. School
    2. Job/Title
    3. Car/Where live

    Yeah, good times.

  • Amol

    Several commenters are using your IQ vs. %Female trend to argue that college men must be “smarter” and have systematically higher scores. That is not correct at all. It is possible to explain this trend without assuming any male-female IQ (or SAT) score difference. Suppose that:

    1. male and female students have the same distribution of SAT-Quantitative (SATQ) scores (i.e., Gaussians with the same mean and variance);

    2. men have a stronger preference for quantitative majors; and

    3. quantitative majors have some sort of admission cutoff (such as the ability to pass a “weeder” course) which is correlated with SATQ scores, so that only high-SATQ students are allowed into these majors.

    As a qualitative example (“toy model”), consider the extreme case that women have no preference for quantitative majors, while men have a very strong preference such that all men with sufficiently high SATQ scores choose quantitative majors. Then quantitative majors will have (i) all high-SATQ men and (ii) only those high-SATQ women who choose these majors. Non-quantitative majors will have (i) those high-SATQ women who choose them and (ii) all low-SATQ men and women. It’s clear in such a scenario that quantitative majors will have both higher average SATQ scores and more men, while non-quantitative majors will have lower SATQ scores and more women, even though men and women have exactly the same distribution of SATQ scores.

    As for why women might have less of a preference for quantitative majors, it could be that a misogynistic society systematically pushes women away from these majors, or that the culture within these majors is more hostile to women. But the upshot is that the trend you found could be explained entirely by differences in preference between men and women (possibly influenced by cultural misogyny) and not differences in intelligence.

    • XorFish

      Fact 1:
      There are 33-50% more women than men on college.
      Fact 2:
      Men are represented more at the extreme ends of the IQ distribution.
      (There are more men than women with an IQ-score below e.g. 80 or above 120)
      Fact 3:
      Only 30% of all men and women graduate.
      Only the part with the highest IQ-score of a demographic graduate.

      With the Assumption, Fact 1 and Fact 3 we can determine that we have to compare the top 24-26% of men and the top 34-36% of women.

      If you add fact 2, it is pretty clear, that the average male college student is more intelligent than the average female student.

      Concidering this, it is important to be aware that that conclusion doesn’t tell us anything about an individual men or women.

  • Geoff

    The first graph is extremely misleading and does not explicitly prove men are smarter than women but instead compares two independent (I think) parameters; Avg. IQ vs. m:f ratio.

    All you can confidently conclude is that e.g. Physics & Astro. have the highest avg. IQ and there’s a trend throughout the disciplines and, that the m:f ratio is lowest for Physics & Astro with a similar trend throughout the disciplines.

    Importantly: the avg. IQ score is discipline dependent and not m:f ratio dependent. It is conceivable for a high percentage female Physics class to exist and you would assume their avg. IQ higher than other disciplines because they take Physics.

    There needs to be a future study conducted on high female percentage classes to verify the aforementioned.

    • Geoff

      Future study: conduct an IQ test for every University Physics class in the country and their m:f ratio. Plot.

  • LOL

    Lots of hurt feelings in here. It’s really as simple as the fact that men as a whole are smarter than women. Didn’t need a scientific study to know that.

    • Dee Doubs

      They are also stupider as a whole if that makes women feel better.

      • Raymond Martin

        That makes no sense. If it were on the whole then there wouldn’t be men above women at the higher end of IQ. But there is, about 3-4 times as many.

        • Dee Doubs

          Alright, you see how you are looking at the higher end of IQ? Ok, cool. Now do the same thing for the bottom end of the chart where men also outstrip women by similar numbers. That’s what I mean by they are both smarter and stupider as a whole.

    • Morgan

      If you were an intelligent person, you wouldn’t have made that remark. :p

  • dave

    And when interpersonal and social intelligence is measured, these charts would be inverted.

    And when linquistic intelligence is measured, these charts would be skewed hard to favor the right middle.

    The charts here are strictly based on test of logical intelligence, but psychology has classically defined nine categories of intelligence.

    What a joke.

    • Dee Doubs

      Wake me up when you find reliable tests for gauging all those categories of intelligence. Then maybe we can start charting those things. In the meantime you are just blowing smoke up people’s asses.

  • D

    Grouping things in such a way somewhat misrepresents the data. For instance, imagine a thought experiment/scenario where there are 3 majors (A, B, C), which have average IQs of 130, 120, and 110, respectively, and have the following percent females: 20%, 50%, 80%. (Giving you a very clear inverse relation). Let’s say the women within each category are always at the top of their major. For instance (chosen to make mental arithmetic easier), class A has female IQ of 160, male IQ of 122.5, class B has female IQ of 150, male IQ of 90, class C has female IQ of 110, male IQ of 60. This means that the average female IQ is 130 and the average male IQ is about 103, within this theoretical system. Such a thought experiment corresponds to the scenario where a woman has to be far better than a man to enter a particular major. For instance, a woman might be more likely to drop out of a technical major if they have a GPA of 3.2, while the “failure threshold” for dropping out might be a 2.7 for a man (I think I read that this has been shown in some engineering schools). Considering that most of the girls in my compsci classes had top marks, I think this is not an unreasonable model…

  • Kevin S. Van Horn

    “Many women tend to avoid mathematics due to social stigma, lack of exposure during childhood, etc.,”

    How do you know this? I keep on seeing people make this claim, but nobody ever backs it up with any evidence.

    I have to wonder how much of it is just that men and women gravitate to different interests. I have two daughters who both completed calculus in high school with A- to A grades. They got plenty of encouragement from me — I really enjoy math, and I have always shared my enthusiasm with my children. But neither of them seems to have any real interest in math or mathematical subjects, and they both chose majors in the humanities. This is, of course, just a single anecdote, but it is representative of my observations over the years.

  • Anon Ymous

    SAT to IQ conversion seems to be off. Here is a good site for SAT to IQ conversions:


    The smartest major, physics and astronomy, have an average SAT score of 1270. This seems plausible. However, your source converts that 1270 to an IQ of 133.

    As the source I provided above shows, the actual conversion of a 1270 SAT score to IQ should be 128-130 depending on the IQ test you use.

    Because this post is about the relative IQs of majors, this doesn’t actually have any impact on your argument. But, it is important to note that it seems there was an error on the part of the source when converting SAT to IQ.

    My suspicion is that it has something to do with how they handled selection bias. Of course, I could be wrong.

    Overall, though, good post.

  • Jim

    It might be OK to believe the first graph. There are objective measures there which we can take as fact (The fields of study and gender balance). That leaves IQ which is a number based upon testing. Perhaps males are simply doing better at the tests and that might be explained by something other than innate difference in intelligence. There might be multiple factors that bias male ability to score better on these tests than females. I propose (and not from a position of PC) that these are natural and understandable. All we need to do is identify and prove their effects. For example (and I believe there are many more) there is an expectation difference imposed upon males that differs from that imposed on females that statistically says “the buck stops here guys”. From childhood, many males and fewer females are socialized to understand their role will be as a provider and not a providee. I contend that the female is much more likely to understand one of her options is to attach to a male for support. The result might be more vigorous preparation by males to conquer what is tested, based upon their perception of having fewer options to serve in a dependent role.

  • http://GonzoEcon.com Tony Lima

    I’d be interested to see the interactive graphs for verbal and quant SAT, too. But you’re probably too busy dealing with comments!

  • Pingback: US College Majors: Average IQ vs. Percent Female. | Deft News()

  • http://www.daviddfriedman.com David Friedman

    Did you actually calculate a correlation coefficient for the verbal score? Eyeballing it, I would have said the correlation is still there, just weaker than on the other two graphs.

    • http://www.randalolson.com Randy Olson

      Yep. The R^2 on the Verbal SAT vs. major’s gender ratio graph is 0.019 — a very weak correlation by any means.

  • Pingback: ¿Qué estudian los hombres y las mujeres de EEUU según su cociente intelectual? | Diego Sánchez de la Cruz()

  • Pingback: Average IQ of students by college major and gender ratio | Randal S. Olson | newstesttest()

  • Camden

    Also, it’s been shown repeatedly that boys are systematically discriminated against by female teachers, which clearly damages their education. When pupils’ work is blind-marked, boys do better than girls in every subject – a lot better in maths and science and a bit better in English and Biology.

  • Peter

    Those charts just look like made by Fox news.

  • Alex

    What about a plot of IQ spread per field. It seems to me that IQ is good at measuring the skills necessary for math-intense sciences, but is it a generally useful measure of “intelligence?” So if we find, overall that the same fields dominate the upper end of the IQ bracket, then the plot vs. gender distribution is pretty moot.

  • Dee Doubs

    > http://blogs.warwick.ac.uk/images/georgeriches/2006/12/04/variance.gif

    Holy crap, it’s almost like if you culled the bottom seventy-five percent of the chart that you’d realize that the average IQ across the entire population isn’t important… only the sizes of the populations that are able to perform at a level necessary for completing a given major.

  • Kennon Gilson

    Provocative article. But Randy, RE: ” The higher the estimated IQ, the more quantitative the major (with Philosophy being an odd exception),…”
    IMO nothing odd there except likely people not being aware of what goes on philosophy or confusing it with religion/people taking philosophy as a ‘hobby’ or self-enrichment second major. Most serious philosophy major programs in the US presume you’re well-versed in both theoretical and applied calculus III, statistics, a systems course, and mathematical proof theory. That’s basically a math/engineering minor.
    When the Libertarians finally convinced the SATS and GRE’s in a letter-writing campaign to separate philosophy majors from religious studies, the GRE verbals and logic of philosophy majors blew everyone out of the water, and their maths are right up there with engineers. No surprise; philosophy majors who don’t go into teaching/scholarship/arts/computers are preferred to go into top law, business, medicine, where the typical IQ is 150+. My understanding is they also tend to be serious multi-majors e.g. philosophy and economics, math, psychology, history.

  • Mo321

    I am a 56 year old electrical engineer my life experience tells me that women are MUCH more emotional than men, women LOVE to feel their emotions, and LOVE work where they can share their emotions with others and others can share their emotions with the woman NOT MATHEMATICS! Men love quantitive subjects. Even in every day conversation men and women talk differently Men will ask quantitive questions – how fast can it go?, how many gigabytes is it? etc Women ask feelings and emotions questions How do you feel?, does he love her?, Which one do you enjoy the most?
    Society wants to convince us that men and women are essentially the same, but my life experience has taught me that men and women are very very different

About this blog

The data visualizations on this blog are the result of my “data tinkering” hobby, where I tackle a new data analysis problem every week. If I find something interesting, I report my findings here to share with the world.

If you would like to use one of my graphs on your website or in a publication, please email me.



Enter your email address to subscribe to this blog and receive notifications of new posts by email.