Isabella vs. Alex in the lab

Girls with “feminine names” are less likely to study advanced math or physics, new research concludes. Abigail, Lauren and Ashley, considered less feminine names by the researchers, are more likely to study high-level math and physics than their sisters with the more feminine names of Anna, Elizabeth and Emma.

The effect is so strong that parents can set twin daughters off on completely different career paths simply by calling them Isabella and Alex, names at either end of the spectrum. A study of 1,000 pairs of sisters in the US found that Alex was twice as likely as her twin to take maths or science at a higher level.

Researcher David Figlio, a University of Florida economics professor, blamed stereotyping.

Figlio pointed to the controversy that arose over the first talking Barbie’s phrase, “math is hard.” “It is a stereotype, and girls with particularly feminine names may feel more pressure to avoid technical subjects,” he said. Not that they were any less capable. When the Isabellas, Annas and Elizabeths took on their tougher-named peers in science, they performed just as well.

Gene Expression questions the sample size. There are a lot fewer girls with “unfeminine” names such as Alex, Abigail and Grace.

I wonder about the decision that Elizabeth is feminine but Ashley is not. If we were comparing “Brandi” to her twin sister “Morgan,” I could see the difference. Is “Emma” more feminine than “Emily?”

About Joanne


  1. Richard Nieporent says:

    If I were an economist I would be embarrassed that someone calling himself an economist could come up with a study as foolish as this. Yes, the size of the population studied is more that suspect. Also, the idea that a name like Ashley is less feminine than Elizabeth is begging the question. Shall we say that by deciding what was and wasn’t a “female” name, all he did was make the data fit his conclusions. It is one thing to claim that people discriminate against females. However, to claim that people discriminate against “female” names is ridiculous. Would you not think that the fact that a teacher is able to see that the student is a female would have an infinitely greater effect on the teacher’s actions than the student’s name?

  2. Twill00 says:

    Wow. Where to begin. To me, Ashley and Abigail are clearly feminine names. The question then becomes, what criteria did these people (“researchers” may be too kind a word) use to determine what might constitute a “feminine” name?

    Okay, I looked at the description of method, and the example ratings, and it seems to be, roughly, that long Latinesque is feminine, saxon is moderate, and Germanic or Nordic is masculine. Them G’s and X’s up your masculinity. Grace gets a .5 –very masculine –rating. Whatever Figlio is testing, it isn’t masculinity or femininity.

    The second question is, when a girl chooses to call herself Alex rather than her given name Alexandria, might that not indicate something about her personality?

  3. Cardinal Fang says:

    As far as I can tell from the article, one set of researchers developed an algorithm to determine how “feminine” a name was, and a second set computed how likely a girl was to study science based on the “femininity” of her name. So the second set of researchers can’t be said to be making the data fit their conclusions.

    Moreover, the name/science researchers studied 1000 pairs of sisters. If they were randomly selected, that is an ample amount to make valid conclusions, especially for a large effect, as this is said to be.

    There could be many other reasons why the study is valid, but those two objections don’t seem sound to me.

  4. Kirk Parker says:

    I wonder about the decision that Elizabeth is feminine but Ashley is not… Is “Emma” more feminine than “Emily?”

    Shhhhhhh! They got the results they were after! Don’t bother them…

  5. Walter E. Wallis says:

    Are we even allowed to talk like this any more? Look what happened to Imus and that college ex-president what’s his name.
    Did anyone ever call our hostess [are we allowed to say hostess?] Jo?
    Men, just shut up, you can’t win.

  6. Tracy W says:

    Alternative explanation – parents who are more likely to have daughters who are likely to study science, mathematics, etc are more likely to give their daughters “masculine” names.

    We know that children get their genes as well as their names from their parents, and in many cases children get their home environment from their parents. The causality is not at all clear here.

  7. Eric Jablow says:

    Wasn’t Ashley a male character in Gone With the Wind?

  8. Ashley is a man’s name. Had been for centuries, and still is in some places (Britain, chiefly Scotland). It is, however, a very feminine name in the United States, and few people have ever met a man named Ashley here in the US. Well, at least not one without a heavy brogue.


  9. Cardinal Fang says:

    Cripes, folks, at least read the description of the study before criticizing it. They compared sisters. They found that when a girl has a more feminine name than her sister, she’s less likely to study science. That can’t be explained by scientifically inclined parents giving their daughters less “feminine” names, as both sisters have the same parents.

  10. Richard Nieporent says:

    I used to wonder how rational people could believe in astrology – that the position of the stars at the time of your birth would cause you to act in a certain manner. Of course my fallacy was in assuming that people were rational. When you do a statistical analysis and get a positive result you must be able to identify a mechanism that would cause that result before you can claim that the result is anything other than a statistical fluke. For example, people have been identifying cancer clusters and attributing it to the presence of a high power voltage lines. The theory is nonsense first because they have cherry picked their results. Other locations with high power voltage lines do not show such an effect, but more importantly because there is no physical mechanism that can cause it to happen. Now some researchers are claiming that cell phone use is causing honeybee colonies to collapse. Could it be that the honeybees are running up too many minutes on their phones and are no longer able to pay the rent on their hives?

    So now getting back to our “scientific” study of female names. One group of researches has independently decided what a “feminine” name is and a second group has found a correlation between these feminine names and the careers that women choose to go into. Okay, they have found a correlation in their data sample, but what is the cause? In other words why is this not a statistical anomaly? Just because a group or researchers has come up with an algorithm for labeling names as feminine or masculine does not make it so. Do you really believe that this name naming algorithm is a scientific fact like gravity and evolution? If not then they are basing the results on bogus data so there is no way that their explanation can be correct. As I indicated before, if the reason that females do not go into the sciences is because society is prejudiced against them, then the fact that one can see that they are female would have a infinitely greater effect than what their first name is. In fact if you believed in the more feminine vs. less feminine theory then the researchers should examine pictures of each of the sisters and determine which one looks more feminine. After all wouldn’t that have a much bigger effect on the teacher than the student’s name?

    As a final word I believe Shakespeare had something to say about this in 1594.

    ‘Tis but thy name that is my enemy;
    Thou art thyself, though not a Montague.
    What’s Montague? it is nor hand, nor foot,
    Nor arm, nor face, nor any other part
    Belonging to a man. O, be some other name!
    What’s in a name? that which we call a rose
    By any other name would smell as sweet;
    So Romeo would, were he not Romeo call’d,
    Retain that dear perfection which he owes
    Without that title. Romeo, doff thy name,
    And for that name which is no part of thee
    Take all myself.

  11. Kirk Parker says:

    Uh, Cardinal–some of us who did read the description are complaining about the completely arbitrary and meaningless classification of names as “more” or “less” feminine.

  12. Maybe they need to do some additional studies – with boys named Sue.

    I dunno. My given name (different from my screenname) is a feminized form of a masculine name, and although I’m a scientist and I like math, I’m a pretty girly girl.

    I think it’s possible there’s SOME effect of parents treating kids differently based on whether they perceive them as feminine girls or less-feminine girls and that could have an effect. But I still think the “masculine female” vs. “feminine female” names designation is pretty arbitrary.

    It might be better to do a longitudinal study and look at girls who were “tomboys” vs. girls who were more “traditionally girly” and see what careers they wind up in. I mean, if people really want to know. I take the attitude that if a person is happy, if they’re good at what they do, if they’re making a contribution, who cares? If this research leads people to push girls that don’t want a career into math and science into it just in the name of quotas, that’s not a good thing.

  13. Cardinal Fang says:

    They found the effect. It’s there (if we are to believe the data).

    If the name classification truly were random, then there wouldn’t be any effect, and the “femininity” of a girl’s name would not be correlated with her propensity to study science and math. But it is. So there is something to explain.

    One way the experiment could be bogus, though, is if the experimenters worked backwards from the data. That is, suppose the researcher looked at all the names of the science sisters and the non-science sisters, and tried to find a definition of name femininity that would cause the result he was looking for. That would be a bad experiment, but sometimes bad papers get accepted.

  14. The effect is there if the study’s been replicated by independent researchers and the results of the study have been seen with different conditions like another country, culture or language. Until then it’s just another attempt to wring a few bucks out of the gender gravy-train:

    Researcher David Figlio, a University of Florida economics professor, blamed stereotyping.

    It’d also be helpful if the study suggested some mechanism by which the effect is generated that’s falsifiable.

  15. JuliaK says:

    The Observer article leads me to believe that David Figlio judged each name’s femininity: “Figlio calculated a linguistic ‘femininity’ score for each name. It was arrived at by using 1,700 letter and sound combinations that could be associated as either female or male and matching them against the names on 1.4 million birth certificates.” Is this method falsifiable? How does one calibrate for a Samantha who chooses to go by Sam? Isn’t this all a judgement call, without any method of measurement? What makes Abigail less feminine than Emma?

    Gene Expression points out the sample size problem. It affects the name frequency, but it also affects the academic tracks available for study. If about 1% of bachelor’s degrees awarded each year are in the field of mathematics, it seems to me that 1,000 twin pairs really isn’t a large enough sample. One math major with a feminine name could have thrown everything out of whack.

  16. Cardinal Fang says:

    If only 1% of girls studied math or science at a high level, that would call the study into question– but we know that’s not the case. Look at how many women are going into medicine these days.

    The way to see if the algorithm’s definition of femininity is the same as people’s would be to test, but presenting different names to people and asking which was more feminine (or by using a more clever test). We don’t know what, if anything, the papers’ authors did in this area- it’ll be interesting to see the real writeup of the study.

    However, it’s important to point out that there is an effect here. That is, there is a difference between the girls with more “feminine” names and their less “feminine” name sisters. (Scare quotes because we don’t know whether “feminine” corresponds to feminine.) We don’t know the explanation for the effect, but there is an effect to be explained.

  17. Yeah, but what that effect is, what it indicates, isn’t clear. It’s entirely plausible that the observed effect is caused by some unidentified factor.

    What’s being assumed here? That this is a basic, human trait? In any culture you look at are the girls with more masculine names going to go into what substitutes for high-level math and physics and the girls with more feminine names going to get their doctorates in bread-making?

    How’s that work in a stone-age culture? How do you determine the masculinity/femininity of names in other languages/cultures? In fact, how defensible is the method used to determine name gender bias in Florida?

  18. Part of what makes this ridiculous is that there’s a lot more to parental name choice than “feminine” vs. “masculine.” Go to
    and plug in Ashley, Isabella, and Elizabeth, for instance. You see that one name was intensely popular twenty years ago but very disfavored now; one was “clasic” but disused twenty years ago but now fantastically popular; and one is “classic” and steadily used (if, like most male and female baby names, slowly dropping in use, mostly because of the proliferation of alternate name choices).

    So we see all sorts of other social factors and (presumably) associations with given names, including class status–a newborn Krystal or Tyffany, most people would correctly guess, is born into a lower income family than a newborn Emma or Ava–and to assume it’s some weirdly determined “femininity level” that’s the most important aspect of a girl’s name makes no sense. There are just so many other factors.

    Frankly, I find it difficult to believe that there are even sisters out there in any significant number named Ashley and Isabella (or Kaylee and Jane, or Michael and Abigail, etc.) Not because these are masculine/feminine pairings, but because they’re wildly out of tune culturally. The family who has used one name for one daughter just isn’t going to have used the other for the other daughter.