Study shows gender bias in science is real. Here’s why it matters.

1. 1. jbairddo 4:09 pm 09/23/2012

   A study which shows people are biased by sex, who knew? But why should this be disturbing? Am I disturbed enough to march on some hallowed institution of science and learning. Am I surprised by studies showing overt sexism and violence in the military? no more than this, and sorry i think the military thing is a lot worse for women. The article suffers because the data are too vague or not detailed. Yeah, women are treated like crap. Big news, women treat other women like crap? Big news. Until one can explain why that happens and more importantly why it does not in the those who didn’t do this, you won’t find a reason for how to change this.

   Link to this

2. 2. blue7053 5:03 pm 09/23/2012

   On the African plains, we developed a division-of-labor scheme that has done well for 200,000 years. It’s called ‘Men’s Work’ and ‘Women’s Work’. So far, so good but it’s under attack because Men’s Work doesn’t get you killed as often as it used to so now there are more applicants.

   First, the division of labor recognized that women have babies and can’t travel well so they stayed home and the men protected the territory. Men become hostile, confrontational, aggressive, in short, all the elements needed to protect the borders of the territory. Women became agreeable, companionable, supportive, sympathetic, in short, all the elements needed to develop a community. Women work better and are safer in groups: they developed instincts of homesickness, loneliness, and fear of isolation.
   Some sub-specie had men who also became lonely and homesick while checking the territory: they failed to thrive.

   Kipling said; “Something hidden, Go and find it.”
   Every man who ever lived felt an urge to go and find his pack after Kipling’s reminder of who we are.
   For the past 10,000 years women have generated the birthday parties, the anniversaries, the box-lunch socials, the funerals, the weddings. Women create the environments in which men may meet without killing each other and despite women’s best efforts, they are not always successful.

   The field of Science is almost entirely composed of ‘multi-dimensional abstractions’. From imaginary numbers to the ‘folding of proteins’, we hold our ideas in our imaginary air, as Sir Isaac Newton described it, ‘constant before us and wait.” While there are women with spatial dimensional ability they are few and experience has taught the lab’s they will probably not generate great success. Nevertheless, the hostile confrontation is not reserved solely for our territorial guards and there are many who respond, “Don’t tell me what I can do. You’re not the boss of me!”

   This confrontational component of our society being addressed herein is a very recent rejection of millennium of developed instinct based on trial and error. In order that everyone may feel good about themselves, we’ve created a unisex test that minimizes the gender differences and many have come to believe that men and women are the same. Larry Summers read 200 years of occupational data and mentioned the gender tendencies and was beset by irate women and their unisex sympathizers. I have no objection to women who want to work their ass off to do what men do naturally and have been doing for thousands upon thousands of years: developing vast complex three dimensional schemes of their imaginings and watching these schemes work out relationships. As Einstein said, “I seldom think in words at all.”

   As I’ve told my women friends for 50 years: if you want to do ‘spatial diminsioning’ as well as men, then do what we did: kill everyone who fails the test! It’s called ‘getting lost’.

   Link to this

3. 3. Jennifer Ouellette 5:35 pm 09/23/2012

   What would we do without the Mansplainers to fill the comment thread with their faux-authoritative ramblings? Seriously, any argument that opens with “On the African plane….” has me simultaneously giggling at the pomposity and calling shenanigans on the supposed “evidence.”

   Honestly, it’s not entirely clear what point blue7053 is attempting to make. But if he is asserting — with a straight face — that the gender discrepancies in the study are due to innate differences in men and women, for cognitive things like spatial reasoning, there are mountains of evidence to the contrary, demonstrating that such differences largely evaporate when cultural factors were removed from the equation. I’d urge him to leave off 50 years’ of lecturing his women friends about his own ignorance and try a bit of Google-fu.

   He can start with the work of Donald Pfaff: http://www.rockefeller.edu/research/faculty/labheads/DonaldPfaff/. See, this isn’t about trying to make the world unisex or deny any and all biological differences between men and women. It’s about making sure that implicit bias and other cultural factors don’t deny women the same opportunities at success as men. It’s people like blue7053 who created that imbalanced world, so small wonder he’s defending it.

   Link to this

4. 4. jctyler 6:19 pm 09/23/2012

   Simply curious: how come gender bias studies are always showing men disadvantaging women?

   Link to this

5. 5. Stephanie Z 7:14 pm 09/23/2012

   jctyler, I’m going to go out on a limb and say it’s because you didn’t read the studies or any of the coverage of the studies. This study, for example, shows that both men and women are biased against women.

   Archimedes, I assume you’re quoting for the irony?

   Link to this

6. 6. Paul_D_Heikkila 7:27 pm 09/23/2012

   The authors of the PNAS study have chosen to present results from one case of a candidate for a low level position with “slightly ambiguous competence.” It would be interesting to see if the same bias appears for well qualified candidates and if the bias appears for more advanced positions. In today’s employment situation I suspect there are more qualified candidates than positions. Neither the male nor the female applicant in the current study would be likely to get the job.

   Link to this

7. 7. Na g n o s t ic 7:41 pm 09/23/2012

   The gender of a scientist should make no difference if the scientist is doing good science.

   Link to this

8. 8. RexScientiarum 9:20 pm 09/23/2012

   I certainly am not going to argue that there is no gender bias in science fields but as a senior biology major chemistry minor in college will say that there is a vast difference in the number women vs the number of men in nearly all of my biology and my chemistry classes with the single exception of my botany/plant biology track classes which happens to be my focus. Full disclosure, I am male.

   Now all this is just from my very limited perspective and experience. As a whole women seem to marginally outperform men across the board. Women get more positions in the professor’s labs as undergraduates, more research grants, and a vast majority of internships at independent labs and field stations. As both I and the study mention there are more women than men in many science majors and perhaps that is the the reason it *appears* more women are successful in their job search. Overall I think women are more likely to stick to or even switch into a science major. In general I feel women tend to be more goal oriented regardless of their field of study so I think I might be somewhat biased against men based on my personal experience.

   I assume that the position in the PNAS study was for a microbiology lab or some related biology lab (if it specifically mentioned this in the study I overlooked it, I only read “in the biological and physical sciences”). Perhaps a chemistry lab, environmental testing or something, whatever. I suspect though that the bias would vary between disciplines.

   Undoubtedly this is bias on my part but without reading any studies I would suspect that bias against women would be greatest in a physics discipline and weakest in a biological discipline, perhaps even reversed. I’m not sure about psychology, probably about the same as biology. My bias here is simply based on the number of women I ‘see’ (hearsay) in each field, I would suspect that the fewer women there are in a particular major or discipline the more bias against women there would be simply because of greater scrutiny (scrutiny inherent with being a minority demographic, as such I expect an eventual development of bias against men in fields where women are becoming the majority).

   Not sure where I am going with all this. I suppose I am trying to say I feel like I am in just as much competition with my female contemporaries as with my male contemporaries when looking for undergraduate research grants, internship positions, jobs, graduate schools, etc. but if this study is representative of the science field as a whole perhaps I should feel more advantaged. In today’s job market though I am pretty sure everyone feels they are at a disadvantage in one way or another regardless of gender, race, or creed. We are all fighting the market.

   Link to this

9. 9. astrokiwi 9:51 pm 09/23/2012

   I’d like to see more discussion of the supporting information in this paper. It seemed to me that the authors clearly stated there that they checked for the effect of every factor that they could eg. research discipline, faculty gender, etc. – except for department (prevented due to privacy concerns) – but a lot of comments have focussed on ‘did the study check for the effect of x’.

   Which thoroughness makes the results…even less cheering.

   Link to this

10. 10. lordsev 10:04 pm 09/23/2012

    2 things:

    1. A similar study (Bertrand and Mullainathan 2002) was done in which job applications were sent either with a black-sounding or a white-sounding name, but which were otherwise identical, and the applications with black-souding names got much lower callback rates.

    2. It’s not clear what kind of discrimination this shows. Clearly it hurts the discriminated group unfairly. However, it might be rational in a situation where the discriminated group has a larger number of underperformers (say, women who’ll end up getting pregnant), but you can’t tell which ones will be underperformers from their resume, so you discount their chances just for belonging to the group. Since this kind of discrimination is rational from the employer’s point of view, an anti-bigotry campaign may not make it go away.

    Incidentally, if this is what’s happening (race/gender sending signals of performance that are not completely covered by a resume), then the two groups you’re comparing aren’t exactly the same from the employer’s point of view, so it’s not a perfect randomized experiment.

    Link to this

11. 11. nikiyaki 11:20 pm 09/23/2012

    I agree with you that the data shows discrimination and as a female, I also would like to see hiring discrimination removed. The only difficulty I can see is the potential pregnancy problem, from the Number 4s. The other article you linked to didn’t really address it at all though, so I’m left feeling like you’ve simply shuffled the issue off without addressing it. As an employee at a small business that has had 4 recent pregnancies amongst its female staff, and being told (by one of the recently pregnant women) “don’t get pregnant, we need you here!” I feel acutely the issue of fertility vs worklife. All these women returned to work, but they did have changing work circumstances to cope with the babies. So what is your response to number 4? The article by your sister seems to imply that the number 4s are really just sexist and therefore their statements can be ignored, but I have seen and felt firsthand the effect of childbearing on women’s worklife.

    Link to this

12. 12. Ilana_Yurkiewicz 11:47 pm 09/23/2012

    @nikiyaki: I linked the piece because I really appreciated the three guidelines she formulated for when employment choices actually cross the line to an “ism” (in this case, sexism). As I touched upon in a previous blog post (Mark Regnerus one), if you want to use group outcomes data to determine when you are justified in discriminating, you would be remiss in stopping at only two groups, or only one group difference. For example: if you want to argue pregnancy, why not look at men’s increased risks of heart disease? Alcoholism? How that decreases work performance?

    While we’re at it, imagine that we learned people who wore glasses were more likely to be better workers than people who did not wear glasses. Would we still make the argument that we should discriminate based on that finding? The thing is, even if group differences *do* exist, it is impossible to predict any given individual’s performance based on them. And, if you wanted to be completely rational about it, that would lead down an impossible road of grouping everyone into multiple categories and analyzing every single difference. The emphasis on pregnancy – but nothing else – seems to selectively point out data that says women will be poorer workers while neglecting all other relevant data that would lead you to the same conclusion about other groups. But, even if all the data were there – the fact that group outcomes data say nothing about individual performance is reason enough to say it’s a bad idea.

    Link to this

13. 13. sharayurkiewicz 12:12 am 09/24/2012

    To further expound upon the pregnancy question, I’d draw a parallel to short men. I think there are a lot of similarities.

    1. Being a short man “handicaps” you the way being a “pregnancy risk” (i.e., a woman) may. In addition to subjective factors like getting less respect, a short man also has increased risk of numerous health issues (depression, heart attack, stroke, Alzheimer’s).

    2. Are these handicaps due to cultural factors or biology? Are short men depressed because they are simply prone to it or are they depressed because they are treated unjustly? Similarly, are there fewer women in positions of power because they are innately less aggressive or because they are socially conditioned to be that way?

    3. There is still debate whether these handicaps affect a short man’s or a woman’s ability to do their jobs effectively.

    4. There is still debate whether we should try to change societal attitudes about the competence of short men and the competence of women, or if it’s simply natural to judge and the injustices are here to stay.

    5. However, we do know from studies that short man (just like women) are already discriminated against in terms of salaries, hireability, etc. (Note: are we trying to explain away this salary discrepancy for short man the way we are trying to do so in women?)

    6. We can easily discern by sight someone’s gender and height (unlike other demographic characteristics that may be able to be hidden).

    So my question is… given all this: do we want to live in a world where the taller man is always hired?

    I think you know my answer.

    Link to this

14. 14. payitfwd 1:36 am 09/24/2012

    As a male who used to work for temp agencies, I can tell you without any doubt whatsoever that the agencies were almost exclusively staffed by women who were very prejudicial against men. Also, attractive people are consistently shown to be given preferential treatment over supposedly homely people. That is not to detract from the validity of this study but sexism runs BOTH ways and there are always going to be improper reasons as to why one job applicant is chosen over another.

    Link to this

15. 15. priddseren 1:40 am 09/24/2012

    What this tells me is the entire Education system controlling science and the people running it are the problem.

    Now get into the real world such as a company with an R&D department and wanting new ideas to make money, they wont care who the idea comes from, unless you actually believe a capitalist company would turn down profit because a woman invented the thing being sold? No way.

    The entire science establishment you are referring too which is run by the Universities and government are where bias occurs all the time. If you are an engineer and you try to publish a paper on particle physics, even though that engineer built particle accelerators, well good luck convincing a bunch of university professors to allow you to publish that one. Just one example.

    In fact, unless you are a product of that good ole boys network of science professors, nothing will get published by you. So yeah bias everywhere but not only against women. Bias is in fact the nature which government controlled entities work by. Maybe, just maybe women are taking an even harder hit in this but it has nothing to do with what the article seems to imply, which is rampant sexism everywhere. It is simply bias the Education/Science industry has against anyone they did not indoctrinate, themselves.

    Link to this

16. 16. jctyler 9:08 am 09/24/2012

    Stephanie,

    It would have been funnier if you had thought for example of a number of (mostly German to my knowledge) studies which show how in kindergarten and primary school boys are greatly disadvantaged in favour of girls and why that is so and the unfortunate social consequences this has especially in cities.

    But let’s be ironic in the classical sense. If it is true that both genders are biased towards one, could one not also at least consider the _possibility_ that they were in fact not biased, but true? If half the witnesses say one thing, they may be wrong, but what if all the witnesses say the same thing?

    Am I being ironic?

    Link to this

17. 17. Nichevo 10:28 am 09/24/2012

    Maybe it isn’t sexism; perhaps women, even those who are equally qualified on paper, on average, don’t perform as well as men.

    I wrote a paper many moons ago about this issue and the differences in the IQ’s of men and women. While the averages are very similar the standard deviations of the males is larger. For this reason one would expect to see a disproportionate ratio of men near the extremes of occupations. It is entirely plausible that women, on average, just don’t perform as well as men in certain areas. So, one is much more likely to hire a male who will receive a future Nobel prize than a female who will. Is it sexist to play the odds, and to base the decision on what people have actually observed, when making hiring decisions?

    Blue’s ramblings aside, it also explains why men occupy the bottom rungs of the ladder as well. I would argue that the ‘disposable male’ phenomenon is much more serious than issues such as these, that feminists worry about. But by the same token, this would explain why these trends exist, and how they aren’t ‘sexist’ per se.

    I should also add that the rebuttal for option #4 is completely lacking. This is disappointing, considering this was such a well written article. I would really like to see this explored further, with proper rigour.

    Link to this

18. 18. Ilana_Yurkiewicz 12:35 pm 09/24/2012

    @jctyler: If you read the first few paragraphs, you will see how differences in outcome cannot prove differences in opportunity. So, if you’re using the schoolchildren example to show gender bias against boys, I’m afraid you don’t have a leg to stand on. If you’re using it to show differences in outcome – well, then that’s just it, and I’m not sure of your point. Sometimes they exist. Sometimes girls as a group fare better. Sometimes boys as a group fare better. The topic at hand, though, is gender discrimination.

    More importantly – in terms of the scientists not being biased, but doing something true: the applications were identical. Let me repeat that: the applications were identical. Identical. Meaning the same. You are saying that it could be true that the same things are different.

    Finally, here is a general note to all commenters. This is Scientific American. This is a forum for rational discourse on content. Commenters are encouraged to disagree, and I welcome a reasonable and informed discussion. But these are the ground rules:
    1) No personal attacks. If you make a personal attack, I will censor you. @thejerk: telling someone to “shut up” counts as a personal attack.
    2) Keep it relevant. It seems one reaction to this piece is “Wow! A piece involving gender. What a perfect opportunity to air all my rants about women/feminism.” That is not relevant. @Archimedes: a long quote from a book that bashes women’s studies in universities counts as irrelevant. @jctyler: linking a 200-page PDF not written in English about early education counts as irrelevant. You can do better than this. Read the piece. Comment on something relevant to the piece. If you think something you’re saying is relevant, show me how.

    Last night, I got wrapped up in a Twitter conversation with other science writers and editors about when to censor comments. For the record, all of them told me to censor far more than I am doing. My policy over the past seven months of blogging has been accepting the vast majority of comments. I do that because I am trusting you all to behave in ways that are appropriate. Please don’t make me regret it.

    Link to this

19. 19. Ilana_Yurkiewicz 12:38 pm 09/24/2012

    @Nichevo, rebuttal for #4 was re-addressed in comments #12 and #13.

    Link to this

20. 20. Frontovik 3:47 pm 09/24/2012

    To rebut this point: “For example: if you want to argue pregnancy, why not look at men’s increased risks of heart disease? Alcoholism? How that decreases work performance?”

    The higher incidents of alcoholism and heart disease in men than women is a trend of the general population, not the selected sample of well-educated young men looking to enter graduate school. It’s generally accepted that there’s a strongly negative correlation between education and alcoholism and a weaker, but still statistically-significant, correlation between health and education. Simply put, alcoholism is not high on the list of concerns for an employer looking for a lab manager.

    Heart disease or other health issues are long-term factors that are also discounted based on the general time horizon the employer is looking the position filled for. If an employer figures that they’ll have this manager for a few years, then they won’t care about his health prospects thirty years down the road.

    However, this is where I realized something interesting that hasn’t really been mentioned: the time horizon of this position is relatively short. The applicant is looking to go onto graduate school, which is in most cases a full-time pursuit.

    How long could be the expected tenure in the position? Six months? A year? That means that the risk of pregnancy causing time off of work (or complete exit from the job market, for that matter) should not affect either the hiring decision or the offered starting salary. (Although, you could still make the case that the mentoring decision is affected by it.)

    I would argue the short-term horizon of employment is what makes this study’s results very interesting. It does make a case for discrimination against women in this case, particularly when it comes to being viewed as hireable.

    There is no reason to hire one candidate over another if the information available makes them exactly the same apart from gender. This does not apply to the starting salary, however.

    One has to wonder how much the differences in starting salary offers are dependent on the ways men and women differ in salary negotiations. I’ve read about a study that finds that men are far more likely to be willing to negotiate their terms with potential employers than females, particularly in first jobs. This may result in providing an employer with an incentive to low-ball a female applicant as opposed to a male.

    The employer is under no obligation to provide a “fair wage” for its employees, but rather get the best deal that he or she can.

    From what’s been presented, it’s hard to disentangle this effect from the dampening affect of a simple unwillingness to hire. I’d be very interested to see if the differences could be separated somehow.

    As for: “The thing is, even if group differences *do* exist, it is impossible to predict any given individual’s performance based on them. And, if you wanted to be completely rational about it, that would lead down an impossible road of grouping everyone into multiple categories and analyzing every single difference.”

    Frankly, employers have to make decisions based on the information that they have given and they’re forced to play the probabilities, which are by necessity aggregate figures.

    The impossible road of breaking everyone down to make note of every single difference is not the logical conclusion of this form of discrimination. If we were able to completely know the abilities of an applicant and translate them into an objective dollar worth, then employers would be able to discriminate perfectly, not discriminate less.

    Overall, I would say the difference in hireability is quite a puzzle for this study given that the typical argument (i.e. women are more likely to exit the workforce to raise children) doesn’t really apply. And, if women are willing to accept a lower starting salary than men, on average, then equal qualifications should lead to *greater* hireability, not less.

    Is the figures for starting salary based on the data for all responses, or only those who respond that they are willing to hire the applicant? I have a feeling that the aggregate data leads to strange results that a more sophisticated comparison would explain.

    Link to this

21. 21. thejerk 4:20 pm 09/24/2012

    Sorry about the “shut up.” I get passionate sometimes when it seems people want to tell me how to run things. I wasn’t telling anybody specific to shut up it was for effect. I was trying to be honest about the reasons males may be hired instead of equally qualified females. Believe it or not, most people who hire wouldn’t be honest about this. When you are in the business of making money, lawsuits go a long way in curving honesty.

    Link to this

22. 22. MickMcC 4:41 pm 09/24/2012

    The results reported in this paper are, sadly, not surprising. There seems to be numerous studies showing that simply changing someone’s name is sufficient to change perceptions of performance. Similar results for racial discrimination have been found. For example, see the studies cited in the brochure at http://wiseli.engr.wisc.edu/docs/BiasBrochure_2ndEd.pdf

    That brochure also suggests that informing people of such bias can help reduce its effects. Therefore, discussion forums like this one are important. However, they are not going to help if they get taken over by evidence-free rants. Therefore, I agree with Ilana Yurkiewicz’s (#18) call for rational discourse; otherwise people (especially those who might benefit from a genuine discussion) will not participate. It is ironic that research on gender bias in science generates a discussion (here and elsewhere) that is filled with so many sexist and evidence-free comments.

    And talking of evidence, people who are saying things along the lines of “perhaps women are not as good at science” should look at the evidence. If they did, they will find research on stereotype threat – for example, lots of studies have shown that women and men perform similarly in challenging mathematical tests under normal circumstances. But if one tells participants beforehand that women perform worse, then the average performance of the women will decline. Similar results occur for racial stereotyping. This (and other research on gender in science) is discussed in a grad seminar class that I help run. You can get more information at http://mickteaching.wordpress.com/2012/09/19/women-in-science/.

    Considering the research on both implicit bias and stereotype threat, if you say “in my experience, Xs are not as good at science as Ys”, it is entirely possible that you are contributing to and influencing that perceived performance, while differences in performance might not exist on a level playing field.

    Link to this

23. 23. thurley 6:07 pm 09/24/2012

    For those discussing discrimination due to possible pregnancy, you do realize that that is gender discrimination right? And that there are prohibitions against that type of discrimination for many companies in the US?

    Link to this

24. 24. Frontovik 10:12 pm 09/24/2012

    @ thurley: You do realize that it’s very difficult to *prove* that a decision not to hire someone due to a likelihood that they’ll quit their job to raise kids, right? You can easily frame the decision by saying one employee indicated that they would be committed to the position for a long-term, while the other one failed to communicate this. Hell, you can lie outright and said one performed well in an interview and one didn’t.

    In fact, by banning that line of questioning, i.e. “Do you plan to leave your job when you start a family,” the government encourages employers to implicitly discriminate based on gender.

    The thing is, I don’t see how one can fault a company who has to make a hiring decision to take potential length of tenure into account and how this does necessarily put women who wish to have families at a disadvantage. Hiring new employees is an extremely costly process for firms and finding temporary replacements that work anywhere near as well as an experienced full-time employee is a dicey proposition at best. (And, if they can, then a firm should wonder if they need a long-term employee to fill the position in the first place. The rise of short-term contract employees is a sign of this way of thinking.)

    Yes, it’s a form of discrimination, but ultimately, any hiring decision is a form of discrimination: an employer favors an Ivy League school over a state school, or an employer favors experience at one company over another, and it’s hard to argue with that.

    But, we also allow an employer to choose a person who makes a better first impression than another, which is a ridiculous metric to evaluate someone on. Easily as ridiculous as a decision based race or gender.

    This leads to all sorts of bigger questions on discrimination.

    Link to this

25. 25. Random Commenter 12:33 am 09/25/2012

    I am surprised by the number of comments that assume some sort of interaction between the hiring manager and the applicant. There was none. So statements such as, “One has to wonder how much the differences in starting salary offers are dependent on the ways men and women differ in salary negotiations,” are irrelevant, because there was no negotiation. The hiring managers received a piece of paper with identical information on it, with the exception of the name. They then answered questions about the person they invented in their minds based on the information on that piece of paper. The answers showed a significant difference in their perceptions of the non-existent applicant based solely on the name – which, in this case, was indicative of gender.

    One thing this suggests to me is that hiring etiquette should require that all resumes be submitted with only a first initial and last name. This would at least give female applicants an equal chance of landing an interview, which would be one step in pushing the “Overton window” toward away from bias.

    In addition, since pointing out the likelihood for bias can reduce the effect of bias, hiring managers should be trained regarding the likelihood.

    The two steps combined could lead to better hiring outcomes in the sciences for women in the US.

    As for those arguing the pregnancy argument as a valid reason for discrimination: to me, the results say something different – it seems to me that we need paid parental leave for both parents. We can do like most civilized nations, and level that corner of the playing field by letting both parents participate in the early days of their children’s lives, rather than punishing the gender that necessarily carries the child to term. Discrimination that leads to reduced wages hurt the family and further hurt the woman when she reaches retirement. If the excuse for wage-reducing discrimination is that she might have a family, then it’s easy to invalidate that excuse by following the path of most other first-world nations.

    Link to this

26. 26. willardthor 3:52 am 09/25/2012

    The blog report neglects to elaborate on the following:

    1. Where (in the world) was the study performed?
    2. Which field / lab was the study performed in?
    3. Which names were used?
    4. Which journal published the result?

    So I looked up the paper, titled “Science faculty’s subtle gender biases favor male students” for details.

    1. United States, only. While the testers made efforts to pick socially-diverse universities, all universities are in the same country, governed by the same government, which does very little to ensure that the sexes are treated equally by the private sector. To avoid the “This woman might get pregnant; let’s not hire her” issue, In Sweden, Norway and Denmark, both parents *share* ~1 year (more in Sweden) of *paid* parental leave. In Norway, mothers must take 9 weeks off (3 before birth, 6 after), while fathers must take 12 weeks off (these are not transferable to the mother, and are lost if not used by the father). In Iceland, each parent gets 180 days, 90 of which are transferable to the other parent. So generalizing the results of this study to science around the world is not justified.

    2. I am not without prejudice; my first thought was “the experiment was probably performed in a biology lab”. Sure enough, the paper says the test was performed in biology, chemistry and physics labs. The paper was not more specific about which labs the application was sent to (I did not see the applications themselves anywhere). However, all these fields have a dominating engineering / experimentation aspect, which cares more about “what works in the real world” and less about “what is mathematically correct”. I dare say that people in such areas are, in general, less objective, and do not apply healthy scepticism as much as, say, certain mathematicians and computer scientists, who only believe facts about their work which are justified by a rigorous proof. In particular, much of this experimental aspect of science applies statistics heavily, prone to inferring wild claims from generated/obtained data (like this article does) (Google “correlation does not imply causation”). I am lead to believe that, since the application was for a lab position, that the people reviewing the application were of an engineering / experimentation part of “science”, and that these people are more inclined to judge based on prejudice and stigma.

    3. John and Jennifer. John, as in “John Doe”, is a typical male name, with no prejudice attached to it (other than “John… how uninteresting”). Jennifer, however, is less typical than Jane, as in “Jane Doe”, and many pop celebrities stupidified by tabloids carry this name.

    4. PNAS. This journal receives funding solely from subscription fees and publication fees, the latter paid by authors. Why is that bad: Such journals might, in hard times, be tempted to publish bad science, since these too yield revenue. PNAS have also been criticised for sending articles to news media before publishing a paper, which then have exagerated the implications of experimental results before scientists have the opportunity to verify them.

    The authors of the study generalize from their experimental results, and the blogger does so even more, taking it as proof that there exists a gender bias in all of science. I do not find these results so conclusive.

    Link to this

27. 27. DrMellio 6:15 am 09/25/2012

    Thanks for this thoughtful discussion. The reasons for my exit from science are now supported by data which, as a scientist, is quite satisfying. I think it is worth noting that any field that excludes or frustrates half it’s talent pool is really just shooting itself in the foot. That talent will invariably find another more welcoming home in which to express itself. Science will be the poorer where other fields are enriched.

    Link to this

28. 28. thurley 6:55 am 09/25/2012

    “But, we also allow an employer to choose a person who makes a better first impression than another, which is a ridiculous metric to evaluate someone on. Easily as ridiculous as a decision based race or gender.”

    Wow! So, not only are you sexist, you’re racist too. Good to know.

    Link to this

29. 29. MickMcC 6:57 am 09/25/2012

    Thanks Random Commenter – good points. The argument along the lines of “but women get pregnant so it is rational to discriminate against them”, is so baseless, it is hard to know where to begin. Comments about how gender-specific tasks were sorted out 200,000 years ago on the plains of Africa are far from insightful (and they worked well for whom?). But, for starters, perhaps we can point out that there are relatively few scientists on the plains of Africa today; and I’m sure lots of things were being done then that we would rather not replicate today. We are no longer cavemen and cavewomen, although I do wonder sometimes…

    But even without equal opportunities for child rearing by women and men, I’d expect that parents are able to sustain a successful and valuable scientific career in the right professional and personal environment. Sure, it is not easy, but there are plenty of simple things that can be done to improve matters. I just don’t buy the argument that motherhood is incompatible with science. There are lots of examples of successful scientists who are mothers, despite entrenched sexism. For example, Jane Elith, one of the most highly-cited authors in ecology, is a mother. Parenthood partly meant she took a different career trajectory, but it is plainly wrong to say that women are a liability because they might become or be a parent.

    The response to problems of mixing science, scientists and parenthood is to change things so it is easier to mix. Saying “it is difficult to mix them, so let’s lump all the problems on the women” is clearly unfair.

    Link to this

30. 30. jctyler 7:04 am 09/25/2012

    I’m somewhat interested in the article because over a period of nearly thirty years I’ve hired dozens of people and sometimes had to fire the one or the other also. I reread the whole article including the comments and a simple question came to my mind:

    How many of you have ever hired anybody? Has the author even had to hire a single person yet? By what criteria? Is she absolutely sure that she would be free of any bias? Would she hire me even if I was perfectly qualified?

    Hiring someone is always subjective. Subjectivity is always biased. Ideally I will hire someone who fits the job best and that I hope to get along with. Everything else is a matter of intelligence and circumstance. If you run a beauty shop catering exclusively to women in a predominantly black muslim community you will simply not hire an obese, jewish-activist white male homosexual, no matter how stylish he is. But you are not allowed to say so unless you want to be flamed by a lot of pseudo-intellectuals and unreal civil libertarians for gay, jewish, fat-people discrimination who never had to run a shop or hire someone. Or who often enough are enormously disliked at their own workplace. If they have one. Because it’s quite often those who don’t get a job because of their unappealing character who scream with the loudest. Would you hire someone who is perfectly qualified but who looks like a pita, smells like a skunk or wears a “I hate society” t-shirt for the interview, regardless of its gender?

    Where you draw the line is the question. How do you keep the pendulum from swinging too far the other way is the problem.

    The problem is real, it is important and the debate has improved a lot of things, but keep it real or else it just becomes another exercice in futility.

    Already, women play the market differently depending on their looks.

    And far more men are hired for LOOKING the job that you would believe.

    Except that the criteria are not the same. See my previous remark on the physical “qualifications” of Wall Street CEOs. That’s bias for you. And no one noticed? Ts…

    Link to this

31. 31. Ilana_Yurkiewicz 9:03 am 09/25/2012

    Thanks for cleaning up the discussion, everyone.

    I think Random Commenter and MickMcC hit the nail on the head about parental leave for both parents. The idea that becoming a mother spells the end of a woman’s career is indeed baseless, and if jobs are inherently structured to destroy families (mothers and fathers too) — the problem lies there, and it’s that we should be looking to change.

    I am also a bit troubled by the comments suggesting that discrimination against women is not a problem to be worked on because discrimination against other groups in hiring exists. That’s like saying we shouldn’t fight cancer because AIDS still exists. Discrimination, overall, is a problem. Just because something exists does not mean it’s *right.* And we have to start somewhere.

    Link to this

32. 32. Martin_Sewell 9:39 am 09/25/2012

    Moss-Racusin, et al. (2012) erroneously interpret the failure of the subjects of their study to commit the base-rate fallacy (ignore background information) as sexism. That the applications were designed to reflect ‘slightly ambiguous competence’ makes the background information (the sex of the applicant) all the more important, and exacerbates the authors’ error. The sex of the applicant contains important information, there are significant sex differences, actually dichotomies, in terms of motivation. In order to attract a high value mate, men have to compete with other men for their rank in the male dominance hierarchy and this translates directly into men contesting each other for positions within organisations. There is no parallel for women. The article by Moss-Racusin, et al. is ill-motivated (in reality, there is no intersexual competition), wrong (the authors commit the base-rate fallacy) and sexist (the authors implicitly deny sex differences).

    Link to this

33. 33. mishie1 1:35 pm 09/25/2012

    Really? People are really going to use #3 as a reason to disregard this information? Figure 2 is showing the difference between the two salaries and the fact that it doesn’t begin at 0 is irrelevant.

    Also, I fall somewhere in between #1 and #2. As a scientist, I think logically and critically about the world that I live in. Did I think gender bias existed? Yes. Am I shocked that it existed to such a degree with *identical* resumes? YES! I’m so grateful to these researchers for giving such compelling evidence for gender bias.

    Link to this

34. 34. evelynjlamb 2:23 pm 09/25/2012

    A few people have hit on the “ambiguous competence” issue. If both resumes had been stellar, would there have been less discrimination? I don’t know. It is *relatively* easy to find the truly outstanding people in a field and give them jobs, whether they are male or female. Emmy Noether, one of the greatest mathematicians of all time, managed to get a doctorate and lecture (under a man’s name) at a time when women were typically prevented from having university positions in Germany because there were people who recognized her as the outstanding talent she was. I am no Emmy Noether, and if I had been in Germany at that time (or anywhere, probably), math would have been closed off to me, while men with similar mathematical abilities would have been able to study and work in that field.
    The top 1% of people will probably be fine one way or another. But it’s also important to ensure that a middle-of-the-pack woman is given the same opportunities as a middle-of-the-pack man. I don’t think that having an “ambiguously competent” candidate is a reason to discredit the study.

    Link to this

35. 35. ancym 3:15 pm 09/25/2012

    Sorry, I know this is beside the point of the article, but the last paragraph left me wondering “huh?”:
    “I once knew of a professor who consistently made eye contact with males when engaging in conversations about science; only when it was pointed out to him did he realize he was doing it, and he was grateful that someone told him so he could change.”

    What am I missing?

    Link to this

36. 36. Geekoid 4:37 pm 09/25/2012

    While I don’t disagree with the study, one study isn’t proof, and you should know better.
    I would argue this study type needs to be broadened before we can start talking in conclusive terms.

    I also think proper hiring practices could clean this up quite a bit.

    Link to this

37. 37. inspiringscience 7:12 pm 09/25/2012

    This reminds me of a paper published in Trends in Ecology & Evolution several years ago. In 2001, the journal “Behavioral Ecology” changed their peer-review policy to be double blind. In other words, the reviewers of a manuscript were no longer given the authors’ names. Amber Budden looked at the proportion of female first-authorships in the journal before and after the change and found an increase of nearly 8%; meanwhile, there was no similar increase in another journal which continued with traditional peer review. The actual paper (http://www.sciencedirect.com/science/article/pii/S0169534707002704) is behind a paywall, but a quick search turned up an old blog post about it: http://blogs.nature.com/peer-to-peer/2008/01/doubleblind_peer_review_reveal.html

    I do agree with the previous commenter who pointed out that this study was based on US universities and so the results should be interpreted in that context. It would be interesting to conduct a similar study in other countries, especially Northern Europe, which is often thought to treat the sexes more equitably.

    It would also be very interesting to see what the results would look like in different fields. Is there a general perception that women are less competent/reliable/etc, or would we find that there are jobs in which men are discriminated against? Perhaps jobs which are traditionally associated with women, such as schoolteacher, nurse or nanny?

    Thanks for your interesting post, Ilana, and particularly for the link to the Implicit Association Tests. People seem loathe to admit that they might have an unconscious bias, so it’s great to have some kind of test.

    Link to this

38. 38. Sam B 11:49 pm 09/25/2012

    I think a few of the people reading this article may have missed the point about what it proves and what it doesn’t prove. Now, I believe the researchers showed pretty clearly that gender bias plays a role in hiring discrimination in U.S. universities in this instance. As such, gender bias is a likely factor to point to when trying to explain the gender gap in science and more research should be done to see if this plays out in more substantial ways.

    However, what this research has not proved is that gender bias is the ONLY factor in explaining the gender gap. There have been many other reasons brought forward that the gender gap exists, and this paper does nothing to discredit them; it just controls for them in order to see if the gender bias exists, which it has been found that it does.

    Some other factors that may also contribute to the gender gap are lifestyle choices (no matter the reason, if a woman chooses of her own volition not to have a career in science, we would expect lower numbers of women in the sciences). And variability of intelligence (see Nichevo’s comments above – some research points to male IQ having a higher variance than female IQ (google it). If this is true, we would expect to see more men at the extreme ranges of IQ even if average male IQ equals average female IQ. The point being here that IQ is correlated with performance in science). If the lifestyle argument holds true, we should look further into the reasons women may not choose a science career. If the variability of intelligence argument holds true, it may be harder to resolve it, except for continuing to provide excellent educational opportunities in science to both sexes. We don’t know for sure which factors are the most important, or even if the two I listed above hold any weight to the gender gap, but research should be to see if this is the case. If this research is not performed, then we are not being truly rigorous.

    My main point is that real life is often too complex to accept just one factor as the main cause for something. The notion of multiple explanatory factors should be taken into account, and we should try to determine which factors have the greatest effect.

    Link to this

39. 39. craig_russell 8:05 pm 09/26/2012

    Please see the following citations for earlier work on 1) bias in rating job performance by race & gender and 2) bias in evaluating the relative worth of jobs:

    1) Hamner, W.C., Kim, J.S., Baird, L., Bigoness, W.J. (1974). Race and sex as determinants of ratings by potential employers in a simulated work-sampling task. Journal of Applied Psychology, 59, 705-711.

    2) Schwab, D.P. & Wichern, D.W. (1983). Systematic bias in job evaluation and market wages: Implications for the comparable worth debate. Journal of Applied Psychology, 68, 60-69.

    I wish there was an easy “fix” for gender and race effects. I suspect as long as there are differences (cultural, gender, etc.), people will find ways to view them positively/negatively in making all sorts of important life decisions. As a middle aged, white male raised in the Midwest, no matter how hard I try I cannot become a young black Hispanic female raised in Miami and view the world with her perspectives. Being aware of the various lens we view the world through may mitigate, though I suspect will never totally remove their effects.

    Link to this

40. 40. Abraham Kuruvilla 10:54 pm 09/26/2012

    The world of literacy is dominated by the left-brain, the male brain. Therefore, as long as research papers or scientific publications are presented in the form of printed words alone a male bias is bound to persist, just as a there was a very strong female bias when oral traditions and pictorial scripts prevailed, in the distant past.

    With the coming of secondary orality; a combination of full motion video, audio and text that is fast overtaking pure literacy there would greater balance between the left and right brain and that would perhaps foster gender equality.

    Link to this

41. 41. wmroche 11:26 pm 09/26/2012

    I just finished reading the book “The Girl with the Dragon Tattoo” by Stieg Larrson and at the beginning of each chapter is a statistic about the startling degree misogyny of many men toward women in Sweden. A study of this type done in Sweden may show similar results in spite of the superior treatment in Sweden of women on maternity leave.

    The original title of the book translated from Swedish is “Men Who Hate Women”.

    Link to this

42. 42. NaLalina 3:29 am 09/27/2012

    My, there’s a lot of Explaining Away the findings and Apologetics of discrimination up in here. First off, just-so explanations are useless but for defending status quo, or stagnation. Secondly, regarding volition, IQ, and some other factors people gave to justify the discrimination the study found (a study based on previous studies also finding implicit biases): all those factors are changeable; see “Flynn effect” and longitudinal studies of IQ for examples. (I would also point to an information theoretical understanding of volition and intelligence, but that’s still in my head as part of the theory of everything macrophysical that I’m still refining.) Thirdly, stupid discrimination in opportunity is discrimination against all of us gaining more from individuals’ works.

    Sam B, it’s not that the persons explaining discrimination don’t get that the article isn’t claiming one factor; it’s that their brains are explaining away problems they don’t want us to care about, like Nichevo implying that stupidity about sexes can’t happen among scientists (or among mathematicians and computer scientists, according to willardthor, as if numbers somehow gives one accurate knowledge on women and men; Mary’s room, anyone?). For several commenters, stupidity in hiring is not a problem if women don’t get hired. They care more about men, specifically themselves, getting hired. Nichevo even throws out that “the ‘disposable male’ phenomenon is much more serious than issues such as these, that feminists worry about.” Nichevo is obviously an anti-feminist male ignorant of the problems feminists address, including males’ problems and economic problems, unhelped by apologetics of status quo and denial of many problems, as if 7 billion humans should focus on “the ‘disposable male’ phenomenon,” while Nichevo probably doesn’t help such males anyways (no, it is not help to fill their heads with rage against feminists or, as some anti-feminist men dispense, tips for abusing women).

    So, Nichevo, willardthor, and several others, I’m calling your brains out on their BS. I don’t write such things to make personal attacks, but to help people become aware of and gain control over their brains carrying out apologetics, BS explaining (observe: http://lesswrong.com/lw/elg/new_study_on_choice_blindness_in_moral_positions/, http://lesswrong.com/lw/e25/bayes_for_schizophrenics_reasoning_in_delusional/).

    Sub-optimal discrimination is essentially drawing on too little raw information, such as in low-effort thinking, which is also a factor for conservatism (http://psp.sagepub.com/content/38/6/808.short?rss=1&amp%3bssource=mfr). The point of studies like the one discussed in the article, indeed the point of science, is to give us information we might not otherwise have known, though some of us could’ve predicted such results based on what we’ve already known, as in response #1 to Carrol’s post. However, the article hardly gives or activates knowledge to fill gaps currently filled by confabulations when people are prompted to explain the results. There are other studies showing examples of where biases come from, like Nalini Ambady’s work involving TV clips. But really, any normal brain should have plenty of data on where biases come from. Even as a child, I could see biases and BS from adults and media. The least we can do is minimize our own BS! Nobody needs you to explain the results (in fact, it’s annoying: http://articles.latimes.com/2008/apr/13/opinion/op-solnit13). Make yourself more useful.

    Link to this

43. 43. JayMan 5:34 pm 09/27/2012

    I have written a response to this on my blog. I invite you to read it, but I ask you to please be sure to read through to the end to see my prescriptions for the problem. Thanks!

    The Leaks in the Pipeline Found? | JayMan’s Blog

    Link to this

44. 44. nkirv 11:05 pm 09/27/2012

    If the skill required is financial management, then at least development agencies know to be biased to favor women: it’s well known that women handle household finances more responsibly than men in development countries, so aid is always channeled to the woman of the household. Men on drink, entertainment, and risky ventures, while the women will support the family and home. From what I’ve seen of among my friends in the U.S., who were never in debt until they married their financial dunce husbands, I’m beginning to think this gender difference is not limited to developing countries.

    Link to this

45. 45. Steve Parkes 5:51 am 09/28/2012

    ancym at 35

    I assume this was as opposed to a lack of eye contact with females. It would have the effect of seeming like he’s engaging with males more. Engaging with one group can give a subtle que that you value one group over the other.

    Link to this

46. 46. artichoke 8:22 am 09/28/2012

    There’s another possible explanation, in addition to the four listed, that could explain why everyone (both men and women) rated the qualifications lower when they were attached to a female name:

    (5) The applicant has been the beneficiary of affirmative action, and therefore she did not have to jump through hoops of the same difficulty all along the way.

    This is an absolutely rational way to interpret data when one knows that affirmative action exists. It certainly exists pervasively with respect to race. It’s less clear with regard to gender. In fact because more girls apply to college than boys, it’s reported that colleges seeking a 50% male class admit boys a little easier than girls.

    But in science and engineering, there’s always female “under-representation” and always someone saying we’ve got to do something about it. There is always a push for affirmative action, to get more females! This has been true during my whole life, ever since I started in college and could observe it.

    I remember my freshman year in engineering school. Usually the class was 20% female, but that year they tried increasing it to 25% female. They admitted an extra 5% of the class as females with somewhat lower standards, to get the percentage up.

    By the end of the second year, just about all of that 5% of extra females had flunked out. And this was a very meritocratic place, if you were clever they didn’t care if you were a small green man from Mars.

    Link to this

47. 47. artichoke 10:43 am 09/28/2012

    Some comments in this thread seem quite un-objective and un-scientific. For example:

    (1) Dismissing the relevance of what actually worked as described on the plains of Africa (well to be precise, the comment-on-comment said “planes” although the original had been “plains”) according to a logical argument. Another in the same vein says “we are no longer cavemen and cavewomen”, again avoiding the logical argument presented, and that we are still men and women. Is it evolution or creation? I don’t know, but it is the way women are and men are.

    Another story from my past, now even further back in the mists of time, but biologically I think our species hasn’t changed in 50 years. Like many here I took the most advanced math that was available in my high school. Year by year the percentage of girls went down, until as a senior in our advanced-placement Calculus BC class there were five boys and one girl. This girl was very bright and ended up going to Radcliffe (Harvard) the next year. But she dropped out of our class back to the Calc. AB class. We missed her. We were nice people. We didn’t even know she was having trouble until she vanished. It wasn’t personal.

    (2) Similarly (and I am trained as an engineer at the undergraduate level, but also in applied mathematics and modeling at the doctoral level) I think that engineering prizes objectivity most highly.

    If the engineering doesn’t work, the bridge falls down. If there’s one thing that is very sure in this world, thank God, our engineering works.

    If the math doesn’t work, one can often tweak the assumptions to make a new pretend-world in which it does, and if a discussion of real-world credibility ever occurs, it’s often a matter of taste and the tweaker can bring argumentative and political pressure to bear.

    (3) As noted by someone else, it’s not clear that gender adds no performance-based information, given the other information that may have been in a resume. There can be differences in the interpretation of results achieved and listed (I noted this: did affirmative action help? hurt?) and the projection of that information to the future (Another contributor noted this: conditional on the information, is one applicant more likely to have more still-latent ability? one aspect of this still unmentioned is that men mature later than women in many ways.)

    Link to this

48. 48. PhysicistChic 11:57 am 09/28/2012

    Articoke- if all the “subpar” women flunked out, how would applications from graduates reflect affirmative action? Your story illustrates that only equally competent men and women graduate and therefore are would be applying to the position.

    From my own experience, there is still a hostility in many science departments toward women. When I was an undergraduate a professor in my first physics course took it upon himself to tell me that the likelihood of me being able to complete a physics degree was equivalent to the likelihood of his 12 year old climbing Everest. This was only the first in a series of unfortunate incidence that made it a very different environment for me than for the men in my department. Although we like to imagine hostilities were left back in the 1950′s, there is a segment of the population who believes women are inherently inferior in the sciences.

    Link to this

49. 49. contrarian246 1:49 pm 09/28/2012

    Apparently no one has told artichoke that the plural of anecdote is not data.

    As long as we’re telling stories to make points, here’s one of mine. When I was in high school, every winner of national science competitions was female. We had two males on our competitive mathematics team, but they consistently scored at the bottom of the group. Same with BC Calculus. Interesting, however, that didn’t stop these boys from applying to places like MIT and ranting against girls when they predictably got rejected. From my experience, the boys seemed to have dramatic differences between their intelligence and their egos. When things went well for them, they took credit, but when things didn’t go well, they blamed everyone else.

    What does this prove? Nothing, actually. I was not so irrational to think that because of these experiences that all males are less competent. Nor would I judge an identical applicant as less competent simply because he was male and reminded me of the many other incompetent and unintelligent men I knew. I don’t discriminate against males in hiring because every person is different, and I judge each one on the content of their work. I can’t imagine doing anything different.

    Link to this

50. 50. artichoke 4:02 pm 09/28/2012

    Indeed anecdotes are not verifiable. I generalize from my years of experience. By the way the engineering school I was describing is MIT.

    Those not flunking out were indeed showing themselves to be good enough to graduate. But it doesn’t mean they were equally represented in the middle of the class or at the top, nor that if they’d admitted an extra 5% of boys they would have mostly flunked out either.

    Nor does it say how many of them majored in the more challenging areas, vs. those who took no hard technical classes beyond the General Institute Requirements. I have the impression that there was a difference here as well.

    It is pretty remarkable to see a flunkout rate 5% (of the whole class size, or 20% of the girls) higher for my class year than for previous ones. It is strong evidence that many girls admitted were barely good enough; they had been “pushing it” already to get 20% of the class to be female. Frankly if you didn’t see this interpretation and thought it meant that the remaining girls were just as good as the boys, that says enough.

    Link to this

51. 51. JayMan 4:14 pm 09/28/2012

    artichoke, very good point. I’ve updated my blog post accordingly.

    Link to this

52. 52. contrarian246 5:14 pm 09/28/2012

    artichoke, I still can’t help but notice how much you’re missing the point. I’m glad you realize anecdotes are not verifiable, because for every one you have I have another one about girls outperforming boys from my Ivy League university. and then i’ll have at least three more about harassment and bullying of girls to counter your story of “we were nice people. It wasn’t personal.” (Not saying your story isn’t true, only that it’s not representative).

    but the point of the article is summarized in the last paragraph: “equally competent women in science are viewed as less competent because of their gender.” Are you sure that didn’t color your impression of the girls you knew? Sure you didn’t judge them harsher, just like all these scientists who were surveyed? The point of the article is that when two people are equal the man is more likely to be thought of as better.

    So what does sharing anecdotes about differences you observed prove? People at the same level should be treated equally. They’re not.

    Link to this

53. 53. artichoke 9:12 am 09/29/2012

    I’m sure there was no harassment or bullying in my HS math class. Absolutely sure of it. We were geeks.

    And I’m absolutely sure I am able to assess talent, at least up to a certain fairly good level. I know there are very bright women in technical areas. I know some of them very well. But they are relatively few and far between.

    It’s rational to take gender into account, when it affects the interpretation of existing data and the projection of those data to predict future performance. This is explained in more detail above. Ms. Yurkiewicz repeats herself in the original blog post to say that the evidence in the study must mean what she says it does — but I disagree for substantive reasons. I know the point she tried to make, but I think it is wrong.

    Indeed people at the same level should be treated equally! Will you join me to say that this means we should get rid of affirmative action based on irrelevant factors and judge people on relevant merit alone? You can still find a diamond-in-the-rough, someone who’s very clever but never got a good education, and hire them over someone else based on that job-relevant judgment.

    Link to this

54. 54. contrarian246 12:29 pm 09/29/2012

    “It’s rational to take gender into account, when it affects the interpretation of existing data and the projection of those data to predict future performance.” Holy extrapolation. There is zero evidence to justify any of this statement. Let’s agree to disagree. And please be reminded once again that discrimination in hiring is illegal.

    Link to this

55. 55. artichoke 12:39 pm 09/29/2012

    Indeed, if there are two equal candidates, discrimination is illegal. We’ll have to disagree on whether the candidates were equal on the evidence provided.

    Link to this

56. 56. Aqua1ung 12:48 pm 09/29/2012

    Gender bias is prevalent in all activities that require intellectual prowess: if you want proof, look no further than the world of chess: the World Chess Champion (http://goo.gl/GYjYX) statistics are overwhelmingly dominated by men. If that doesn’t prove bias, I don’t know what will!

    Link to this

57. 57. JeJblog 9:12 pm 09/29/2012

    I posted this comment on the blog for Jayman, but I thought it had some relevance in response to comments here:

    I’m going to focus on the one (but very big) aspect of your idea about why females don’t get hired-the idea of loading the odds off getting a highly intelligent, high performing employee by selecting the male based on IQ. First of all, IQ, like BMI, should be used sparingly and never in an argument as to why group x is better or worse than group y. White supremacists have been using that old trick for ages as a reason why white people are superior. Besides being Euro-centric, the IQ test also falls into the “stereotype threat” which cannot be teased apart from the raw scores, as well as the “peer group culture” aspect. So if minorities (historically, the black minority) and females consistently have a difficult time scoring in the very top percentiles, is that an inherent flaw in their intelligence, or is it that they are not performing optimally because of the “stereotype threat” or “peer group culture”? So are you actually selecting for the more likely high performing individual if you pick a white male, or are you perpetuating the “stereotype threat” (in particular) by continuing to perpetuate that the white male will be the most successful?

    Link to this

58. 58. JeJblog 9:16 pm 09/29/2012

    In an ideal world we would be able to hire regardless of race, gender identity, sexual orientation, religion, age etc. by only considering the objective assessment of qualification. But hiring without making efforts to diversify workplaces assumes that everyone who applied is applying from an equally advantageous position. We know that females, in particular, seem to be vulnerable to underperforming when they are expected to, so are their qualifications equal to a male that may have been less influenced by others expectations?

    Link to this

59. 59. Aqua1ung 1:03 am 10/1/2012

    @JeJblog: “hiring without [...] assumes that everyone who applied is applying from an equally advantageous position” Yeah, so why not start from Adam and Eve, right? You want to make things right? Then start from Adam and Eve, otherwise you just run the risk of hurting (other) innocents in the process. At any rate, operating such sweeping generalizations such as “we have to choose women over men because women have been oppressed” is definitely the *wrong* way to go: the assumption here is that *all* women were disadvantaged, while *all* men were advantaged. This is why one has to qualify the premise with “statistically,” as in “women were *statistically* opressed/disadvantaged, and men have *statistically* been in an advantageous position,” only this one can in no way justify affirmative action policies. Liberals, of course, know that, and that’s why they’re very elusive in this regard. Liberal ideology can certainly tolerate some “collateral damage” (read “the life of some innocent white males”): that’s the price we have to pay for “Diversity,” the liberal Holy Grail. After all, two wrongs make a right, right?

    Link to this

60. 60. JayMan 10:00 am 10/1/2012

    Here was my response to JeJblog:

    First of all, IQ, like BMI, should be used sparingly and never in an argument as to why group x is better or worse than group y.

    Apparently you haven’t read through my blog. First, welcome! Second, I would recommend that you do…IQ is the largest factor (but not the only factor) that differentiates the relative success of groups.

    Besides being Euro-centric

    That’s a myth which is part of the standard package of disinformation about IQ tests. Consider the fact that Blacks, for example, do even worse on culture-fair IQ tests (such as non-verbal tests like the Raven’s Progressive Matrices).

    the IQ test also falls into the “stereotype threat” which cannot be teased apart from the raw scores, as well as the “peer group culture” aspect. So if minorities (historically, the black minority) and females consistently have a difficult time scoring in the very top percentiles, is that an inherent flaw in their intelligence, or is it that they are not performing optimally because of the “stereotype threat” or “peer group culture”?

    As I noted in this post, stereotype threat cannot be responsible for the lowered scores of some groups with respect to Whites. If it were, and hence underestimating the “true” intelligence of individuals from lower scoring groups, then these individuals would perform better in the real world than their test scores would indicate. But in fact, the exact opposite happens: IQ tests overpredict non-Asian minority performance. Of course, then there is the fact that other groups, such as East Asians and Ashkenazi Jews score better than White gentiles.
    

    Also, your comment that more men have autism spectrum characteristics is flawed. The autism spectrum disorder characteristics were based exclusively on studying males. In females it often is hard to diagnose because they tend to have greater mimicry skills throughout toddler hood and early child hood. Females also have different manifestations, like really high incidences of eating disorders for example. It will be interesting in the future to see the diagnostic criteria change for gender specificity (including intersex) for early childhood.

    So you’re saying that the disorder is the same because the traits are different?? :\ While it will be indeed interesting to see future research on the autism spectrum (and all human psychology), I’m not denying that autism exists in females and that perhaps it is even under recognized. That said, the traits that are relevant (classic autie/Asperger’s-type) are clearly more prevalent in males.

    Link to this

61. 61. contrarian246 12:53 pm 10/1/2012

    @aqua1ung: so, it’s fine going back to the cavemen to say men are better than women, and it’s fine to make statistical arguments (untrue, can I just point out?) that men are more competent than women for hiring purposes — but it’s not fine to do either of these things to make it right? Without even getting into where I stand on this issue, I do want to say thanks for clarifying: apparently, we can use statistical arguments when they favor men, but we must reject them when they favor women.

    Link to this

62. 62. Aqua1ung 8:26 pm 10/1/2012

    @contrarian246: Sorry, I do not understand your argument, and I’d speculate that you do not understand mine either. Let’s try this:

    1. AA policies rely on sweeping generalizations such as “all women have been oppressed, and all men have enjoyed a privileged status”;
    2. Such sweeping generalizations are false (obviously);
    3. Ergo: AA policies are wrong: they are just discrimination in sheep’s clothing;
    4. Corollary: Enforcing AA policies will hurt innocents (but that’s a small price to pay for “progress,” a small price to pay for advancing the “glorious” Big Picture, right?)

    Link to this

63. 63. Bora Zivkovic 1:01 am 10/4/2012

    Hello,

    This is SA Blogs Editor speaking.

    I heard some people have been complaining about having their hateful rants deleted. My response: tough.

    Commenting is a privilege, not a right. The Free Speech clause that Americans cherish does not mean that every online space is a free-speech space for everyone in the world. This is a personal blog – this is Ilana’s living room so she gets to decide what kind of tone the discussions here will be. If unhappy – start your own blog.

    Well moderated communities have the best discussions. 90% of the people do not comment. They are even less likely to comment if they see that the existing commenters are nasty – why bother mixing it up with such people and getting drowned in the noise? But when the comment thread is cleaned up, and actively and obviously so, more people are willing to contribute intelligent discourse – they see that the blog owner is in charge and will defend them.

    SciAm has had commenting for almost a decade now, so there are many people with registrations. Some of them are here specifically to push their agendas, e.g., anti-vaxx or GW denialism, because this is a prominent science site. Until pretty recently we did not have good moderation practices, so these people felt their oats and felt freedom to troll to their hearts’ delight. It is hard work to reverse this quickly, but bloggers are good at that, so a lot of putting people in place happens on blogs, rather than on news articles. Gradually, the commenting community here will become more constructive – and I hope you all help whenever you can.

    I like to keep my own commenters a little on the edge – no clear rules of moderation, which means I can be capricious. So the rule is: don’t make me mad. Which means – start out gentle and see how far I’ll let you go. I also have different criteria depending on the topic of the post (allowing more heated stuff on some posts than the others), and depending on the blog (remember I run something like 5-6 blogs on the network myself – I am much more likely to be protective of Guest Blog authors, or the students on SA Incubator, than on my own personal blog here).

    Sometimes I invoke the rule of “three strikes and out”. When confronted with this rule, some initially nasty commenters disappear, but others say sorry and continue being good contributors. This seems to work on some people – I have gained, through this approach, some very good commenters (though often disagreeing with me on many issues due to different politics etc) who were uniformly banned as trolls from many other science blogs.

    So, different strategies work for different people – some appreciate getting a nice welcome and a gentle reminder of the standards of my blog, while others understand only the language of force. But it is essential for me to always start the process by reminding myself that even the nastiest commenter is a human being, deserving at least some respect and a second chance (which they have a right to blow). And more contentious the comment thread, more important it is for me to be there, reply, show that I am watching, reading and am in control of my own blog space.

    Yet even I had to go to Twitter to ask for help on a couple of occasions when I got avalanches of trolls – it works. Your online friends will come if you call them.

    When it comes to coming to help other bloggers on the network, I am a little hesitant. It is like having a party and then some neighbors call in the cops because it’s noisy. After the cops arrive, the party is essentially over. As the Editor here, I feel I can be perceived as that cop, and I don’t want to stop the discussion and make everyone too nervous to continue commenting. But I have done this on occasion – either when it was obvious that the cop was needed, or in case of some very new/young bloggers, being unsure how to deal with trolls. Most of the SciAm bloggers are quite experienced and savvy, but not all are and I need to take special care of the people who may not be able to defend themselves as well. But call me if you need me, any time.

    So, just to reiterate. This is Ilana’s blog. She sets the rules. She enforces them. If you write nasty stuff and she deletes them, tough luck. Rethink your approach to life.

    Link to this

64. 64. tefish 7:04 pm 10/16/2012

    “What would we do without the Mansplainers to fill the comment thread with their faux-authoritative ramblings?” Thank you, @Jennifer Oullette! My thoughts exactly. How ignorant we women would be without all-knowing males to put everything in perspective for us via comment threads.

    @Blue7054 subscribes to the Larry Summers argument that women must work harder to overcome some inherent gender-based visual-spatial deficiency. He takes the latter assumption for granted based on biased spatial problem-solving studies from a decade ago (which have long been debunked, by the way). While it would be easy to point out the methodological flaws that favored males as well as the cultural factors that inhibited female performance on these tests, I have learned that it is more effective to argue with anecdotes. Like many of my women colleagues, I (a woman) entered the field of engineering because I happen to have strong spatial capacities, and therefore I never had to work particularly hard in engineering school. In fact, most of my income as a student came from tutoring males in my class who struggled with the material.

    Unfortunately, cultural biases influence scientific objectivity as much as they influence the hiring process within the field. To a hammer, everything is a nail: to a male researcher looking to find overwhelming differences between male and female cognitive abilities, they will appear. As with virtually all studies, data are regrouped in such a way that supports the author’s hypothesis; in rarer instances of scientific integrity and honesty, the data are framed to directly disprove the author’s hypothesis. Both instances involve some form of manipulation of the data. Conclusions alone can be misleading, to say the least.

    If sexists want to continue subscribing to the pseudo-science which touts that women are deficient in their capacity for visual-spatial problem-solving (or any sort of mathematical or scientific discourse), then logically they must also subscribe to the parallel pseudo-science — conducted within the same studies — that touts that men are deficient in their facilities with language. If women have to work double-time to produce similar results to their male colleagues in the sciences, then male writers must be working their asses off to produce any work at all. With this inherent deficit, it’s amazing that we have so many bodies of literature from male authors. How hard Fitzgerald and Hemmingway must have worked to overcome their verbal-lingual shortcomings! Fortunately, women in literature are not using flawed studies to blast men in their field; men in science should show women in their field that same courtesy.

    Link to this

65. 65. keppelc 5:01 pm 10/24/2012

    I am curious if this study actually indicates that science is in fact doing really WELL? Perhaps science is approaching equality of opportunity faster than most fields? What I mean is, it would be interesting to see the results from a similar study looking at applicants for, say, a law clerk or account manager position? Extending this thought and in reverse, what about for an entry-level nurse or kindergarten teacher? As someone else noted, it would also be really interesting to see what happens for a senior-level position, again both inside and outside of science. This article is fascinating and a nice shot, but it doesn’t really give the evidence to prove if gender bias in science is in fact on the (relative) mend or not – it might hopefully be!

    Link to this

66. 66. defChildBen 11:43 am 11/1/2012

    Unless I’m reading it wrong, it appears from the supplemental material that the study only used two names, John and Jennifer. The authors of the study probably should read this: http://www.livescience.com/14394-educational-achievement.html .

    Names alone are highly correlated with academic achievement expectations. John ranks as one of the highest. The very highest is actually Katherine, a girl’s name. Jennifer is not ranked highly. This must account for at least some of the discrepancy here.

    Link to this

67. 67. Yesyou 6:48 pm 11/5/2012

    The sad truth is this: So long as women can get pregnant, they will be viewed as unreliable and incompetent in the workplace. I have read more than a few stories of companies that fire women who get pregnant then leave work to give birth. It would take a lot for a woman who has been permanently sterilized – say Essure or Adiana – to prove to her employer that she is competent and can do the job without pregnancy and children interfering.

    Link to this

68. 68. Spaniard 1:11 am 11/13/2012

    @ Random Commentator in reply to

    “As for those arguing the pregnancy argument as a valid reason for discrimination: to me, the results say something different – it seems to me that we need paid parental leave for both parents. We can do like most civilized nations, and level that corner of the playing field by letting both parents participate in the early days of their children’s lives, rather than punishing the gender that necessarily carries the child to term. Discrimination that leads to reduced wages hurt the family and further hurt the woman when she reaches retirement. If the excuse for wage-reducing discrimination is that she might have a family, then it’s easy to invalidate that excuse by following the path of most other first-world nations.”

    A really good example at nipping this in the bud could be found if you or anyone on this blog had some military experience. The Marine Corps allows both the mother and father to take a leave of absence, be it two different types. The mother would take maternity leave (paid in full, obviously, she signed a contract and the gov’t is paying) and the father would take PTAD (permissive temporary additional duty.) The mother has 6 weeks off, away from work with nothing to worry about and a job to come back to and the father gets 10 days (feeble in comparison but for obvious reasons; he didn’t have a child recently.)
    I get it! How likely is it that a for-profit company or science lab that probably needs all of its scientist present will agree, contractually, to pay in full, maternal leave for 6 weeks? Not likely. Here’s the point, there was a time when females weren’t allowed in the services. There was a time when the services didn’t offer the aforementioned amenities to new parents. The Marine Corps has evolved, the Uniformed Services as a whole have evolved. There are still fewer female generals in the marine corps, two to be exact. But that’s one more than last year.

    @Yesyou – I’ve been under the command of female officers in the marine corps that have proved reliable and still very competent even after pregnancy. Might do you some good to be under the command of one them because I can guarantee that you won’t see woman or child-bearer, you’ll see a Marine and a damn good one at that too.

    Link to this

69. 69. julia-scientificamerican 5:36 pm 11/25/2012

    @defChildBen, I’ve read your link, you’ve missed the point. They were not ranking the educational potential, they were rating “existing” educational credentials, which were identical. Furthermore, in this experiment people rating were scientists, not students. Furthermore, there was no name Jennifer in your study (so, how can you say “Jennifer” was not ranked highly if she’s not even there?).

    Furthermore, I’m quite sure that the researchers didn’t discriminate the scientists by the name and culture. Foreign born scientists (high % in research intensive schools), and/or those whose culture doesn’t include English names, might have no idea about “maternal education” of a certain American name (they might even perceive it totally in reverse). But they would recognize the gender.

    Hiring process is undeniably subjective, I would even ask a question if it’s created this way intentionally. Students go to university and their success is measured according to various marking schemes. Then they apply for jobs, but there is no any measuring tool anymore, other then quasi objective hiring and performance evaluation, far from measuring anything (like in Science itself) but rather “assigning” numbers from the top of the head. So, people who don’t like to hire somebody based on their demographics, don’t have to say so, but would rather refer to insufficient qualifications. Very convenient.

    Link to this

70. 70. julia-scientificamerican 6:40 pm 11/25/2012

    One ambiguity with the names though is that somebody called Jennifer would be much more likely then John to be perceived Asian (hence, she might be perceived stereotypically BETTER at Science then John). But assigning them the same last name would effectively eliminate this ambiguity.

    Link to this

71. 71. jaket kulit 12:58 am 11/29/2012

    I think this is one of the most significant information for me. And i’m glad reading your article. But should remark on few general things, The site style is wonderful, the articles is really nice : D. Good job, cheers!

    Link to this

72. 72. Invader Xan 5:20 am 08/28/2013

    I’ve actually met a couple of 4′s. And had some quite heated arguments with them while attempting to explain that their opinions are disgusting.

    Link to this

73. 73. NaomiAlderman 8:34 am 08/29/2013

    Hahaha, all these comments going “it’s because women get pregnant! That’s why no one will hire them!”

    Guys, if this were all that’s going on here, there might be a difference in hireability but not in *perceived competence* in resumes identical other than the gender of the name. It would be a very interesting finding if perceived competence were almost identical, but hireability showed a large difference – that would indeed suggest that a lot of managers were going “hmm, she’s clearly a capable candidate but… I’m worried about what she’s going to get in that womb of hers”. And indeed, that might then be solved by social policy giving men more freedom to stay home with their small kids.

    But. This is not what we find. What we find is: identical resumes are perceived as 10.7% less competent if they have a woman’s name on, rather than a man’s. This means that people – men and women – look at a woman’s name on a resume and instantly downgrade their idea of that person’s competence.

    Now, the difference in hireability is indeed larger – around 22% – which suggests that some of the hiring effect comes from concerns other than competence, the worry about childcare responsibilities might be part of that, more research would be interesting. But a solid half of this is pure prejudice, with people consistently thinking that women are less competent than men.

    For another interesting angle on this, take a look at the studies of “blind” auditions for orchestras. (Spoiler: once auditioners can’t see what gender the players are, female musicians suddenly get hired more: http://www.princeton.edu/main/news/archive/A94/90/73G00/ )

    Link to this

74. 74. Makkabee 3:53 pm 01/10/2014

    “I’m sure there was no harassment or bullying in my HS math class. Absolutely sure of it. We were geeks.”

    This statement alone is enough to destroy artichoke’s credibility. Bullying and sexism are, sadly, well-documented and far too common in the geek community, so the argument that “we were geeks” proves there was no harassment going on demonstrates that the poster is, at best, not a keen observer.

    Link to this

75. 75. AndrewM 4:02 am 04/16/2014

    Just looking at the length of the comments demonstrates that Ilana has done a great job here. The study cited is so elegant, one would have to be deluded to ignore the results. The uninformed backlash, especially those speckled with eugenic and near-phrenological claptrap is so defensive in tone, it is hard to take seriously, but we really need to, since these responses illustrate the abundant varieties of sophism and prejudice behind the numbers. Another study should be run comparing the misinformation, trolling and tone of comments below this article verses those of an identical article with a male’s name appearing as the author.

    Link to this

76. 76. chainlinkc 6:20 am 04/16/2014

    I am woman with a masters degree in Engineering. I have worked in the USA, UK, and Norway.
    While Norway isn’t perfect, and still suffers from some inequalities, they have come much, much further than the USA and UK. And guess what? they have much higher numbers of women employed in STEM fields.

    Maternity leave is only a small part of the picture. Firstly, in Scandinavia, men are as likely as women to take time off for children. Because parental leave is shared or split, as the parents see fit.

    Secondly, girls and boys alike are encouraged to achieve their potential and pursue their interests in education. Cultural biases haven’t been eliminated, but girls and boys score more similarly in STEM subjects on standardized tests. Formerly communist countries are even better in that regard (also numbers of women working in engineering). One could suppose that egalitarianism has some advantages for women )

    Anecdotally, I work in an office (R&D, public sector) where about 40% of the technical staff are women. I experience much less sexism here than I did in similar jobs in the US or UK, from not having to *prove* over and over again that I am competent, to not being subjected to sexist jokes on a daily basis.

    p.s. the economy is pretty good in Scandinavia, too. Maybe there’s a reason http://www.nytimes.com/2011/06/29/world/europe/29iht-letter29.html?pagewanted=all

    Link to this

77. 77. gracehudak 9:29 am 06/24/2014

    Hello! I just want to give a huge thumbs up for the great info you have here on this blog. I will be coming back aimed at your website for additional soon. really great analysis , this is shows quality of the scientificamerican ..

    Link to this

Add a Comment
You must sign in or register as a ScientificAmerican.com member to submit a comment.