Monday, January 21, 2008

Bias Against Women?

 
GrrlScientist claims that women scientists publish fewer papers than men. She then goes on to offer an explanation [Women, Science and Writing].
The fact is that female scientists do not publish as often as male scientists. Why? Some people have told me that women do not produce scientific results that are of the same high quality as those produced by men (nor do they write life science blogs as well as men, apparently) and that male reviewers can readily recognize when a woman is the lead (or sole) author of an article because "women do science differently from men." Basically, science is still a very sexist community where its female practitioners publish less frequently than men at least partially because of the peer-review system that is in place. I think the commonly used single-blind peer review process is biased against papers whose lead (or sole) author is female, just as the field of science is biased against women in general.
She discusses a scientific study that supports the claim that reviewers are biased against women.

This does not sound right to me. There may be all kinds of reasons why women don't publish as much as men (if it's true) but I doubt very much that reviewers are more likely to reject a paper by a women scientist. For one thing, lots of reviewers are women. For another, why in the world would a male reviewer reject a paper just because the work is done by a female scientist? Do you think this is common practice among the scientists you know? Do you really think that in 2008 there are significant numbers of male scientists who are that biased against women? I don't.

UPDATE: Skepchick joins in with [Are women discriminated against in science?]. Let's not forget that this is the important question. There's no denying that some scientists are sexist, just as there's no denying that some are racist and homophobic. Some of them are even (gasp!) Christians! Does that justify saying that science is racist or science is religious? Of course not. What we should be trying to do is decide whether sexism is rampant among scientists or not. If it is, then we can justify saying that women are discriminated against in science even though we might hesitate to say that gays are discriminated against in science in spite of the fact that some scientists are homophobic.


35 comments:

  1. Could it be (i'm 100% speculating here, no data) that female scientists are on average younger and that there is a bias against younger researchers that would show up in such studies?

    If yes this would all resolve itself in a few more years.

    ReplyDelete
  2. Perhaps Larry could give us his reasons (other than accusations of sex bias in science 'not sounding right') for not taking the results of the paper that was cited seriously? They seem at least worthy of that to me.

    Why not just institute double-blind peer review anyway, to save the arguments?

    ReplyDelete
  3. As a female in science, I can 100% say that sexism is alive and kicking, and it doesn't surprise me at all that this is the case. This is not a new finding even, studies have been done in the past and have consistently shown that women fair worse when the reviewer knows their sex.

    Most sexism I deal with is the covert kind, although as an undergrad I was told twice by two different male advisors to just find a husband and be a "housewife astronomer."

    The covert kind is a bit more evil, it's not so easy to identify and laugh off, and the offender certainly doesn't recognize that they are biased. I have worked with many men who have subtle but clear unconscious biases against women. My (and J's former) astronomy graduate advisor is a perfect example of this type of guy. You should chat with her if you are skeptical of sexism in science.

    ReplyDelete
  4. There is, of course, the famous "McKay" study done back in the 80s:

    Association for Women in Mathematics

    It would be interesting to run this again today, to see if things have really changed at all in the last 20 or so years.

    I personally would bet on "some, but not as much as we'd hoped".

    ReplyDelete
  5. An argument from personal incredulity, with no effort to seek data (which does exist, if Larry had sufficient mental agility to employ Teh Google). Outstanding. That'll show those whiny bitches.

    ReplyDelete
  6. You might also want to read CHRISTINE WENNERÃ…S & AGNES Nature 1997, pp. 341-343 (doi:10.1038/387341a0). Darach Watson, Anja C. Andersen and Jens Hjorth have a correspondence in Nature in 2005 that they would like to do a similar study this time in regards to the EURI awards since on the surface it appears there is a bias against women. In other words they want to do the research to test the hypothesis. The ESF would not let the study take place, hence the correspondence in Nautre.

    And then there is the Martell article in the Journal of Applied Social Pyschology (1991) (doi:10.1111/j.1559-1816.1991.tb00515.x) who looked at sex bias and found that "Only when subjects were able to carefully allocate all of their attentional resources did sex bias in work performance ratings abate. "

    Let face it, professors have a lot on their plates. It is not crazy to speculate sex bias influences reviews of articles.

    ReplyDelete
  7. Bias in peer review? Gee, that sounds rather familiar....

    ReplyDelete
  8. ian b gibson,

    Perhaps Larry could give us his reasons (other than accusations of sex bias in science 'not sounding right') for not taking the results of the paper that was cited seriously?

    Because,

    (a) There could be other explanations of the data.

    (b) You don't make accusations of rampant sexism based on a single study of a few hundred papers in Behaviorial Biology

    (c) The accusation doesn't jibe with my personal experience and observations over the past 40 years.

    (d) The accusations don't make rational sense.

    Why not just institute double-blind peer review anyway, to save the arguments?

    Because I prefer the exact opposite strategy. I prefer that reviewers take responsibility for their reviews by identifying themselves.

    Secrecy and science are not comfortable with one another.

    ReplyDelete
  9. george smiley says,

    An argument from personal incredulity, with no effort to seek data (which does exist, if Larry had sufficient mental agility to employ Teh Google). Outstanding. That'll show those whiny bitches.

    This is exactly the kind or response I was anticipating.

    Apparently it's okay to be skeptical of papers that dismiss junk DNA. It seems to be acceptable skepticism to question those scientific studies that find support for the Three Domain Hypothesis. It's all right to question all those articles that say race is a myth. One can even be admired for dismissing hundreds of papers in evolutionary psychology.

    But, is it okay to express skepticism about a few papers that show a bias against women in reviews of scientific publications? Is it permissible to challenge someone who says, "science is still a very sexist community"? No it isn't. Not in today's politically correct society.

    How in the world are scientists ever going to have a serious discussion about this issue if we begin with the assumption that one side is right and anyone who disagrees is sexist? Is that how real scientists are supposed to behave?

    George, please answer a question. Do you believe everything you read in the scientific literature?

    ReplyDelete
  10. How in the world are scientists ever going to have a serious discussion about this issue if we begin with the assumption that one side is right and anyone who disagrees is sexist? Is that how real scientists are supposed to behave?
    ********************
    Of course Larry you took the opposite explanation based on anecdote. You dismiss the other hypothesis because it does not jibe with your personal view of the world. Not very scientific. And then you frame the other side as being unscientific and "PC". PC taps into certain frames in society.

    ReplyDelete
  11. melissa says,

    As a female in science, I can 100% say that sexism is alive and kicking, and it doesn't surprise me at all that this is the case.

    I shouldn't have to say this, Melissa, but I am not denying that sexism exists. There's a big difference between recognizing that some scientists are sexist and making blanket accusations about all of science. Don't tar every scientist with the same brush. That's not rational. It means that you are insulting the very allies you need in order to make changes.

    Yes, I know who you're talking about and in additional to being sexist he has many other faults. Do you think every male scientist is a total jerk who cheats on their wives and has an extraordinary fondness for male undergraduates?

    I think there could easily be differences between the disciplines as well, but even then, I would hesitate to say that physics (or whatever) is a sexist discipline. That would do a great injustice to many physicists.

    If you turn this into a men vs. women issue then it will never be fixed. Why not try and identify those individuals who are causing the problem?

    Here's an idea. Let's go back and examine the study in Behaviorial Biology. If it turns out that sexism can be demonstrated, then we should be capable of picking out those individual reviewers who rejected papers authored by women before the double blind study was initiated. Get rid of the bad apples but don't denigrate all male scientists. We have a word for that kind of stereotyping, don't we?

    ReplyDelete
  12. Can you propose another explanation for why the simple act of double-blinding produced such a dramatic difference? And remember that both pre- and post-double-blinding rates come from several years of the journal, which ought to smooth out any random noise.

    If you want a lot more data, here is a meta-analysis of no fewer than 21 studies on gender differences in grant peer reviews. As the authors say quite forthrightly, establishing differences cannot prove the existence of bias. But it seems to me that when the differences are as consistent and widespread as they appear to be, one has to at least seriously consider that as a possibility. Furthermore, since good brains come in bodies of both sexes, it would a waste of talent and therefore a disservice to science not to get serious about identifying and eliminating the reasons, whether they be bias or other factors (or most likely, a combination of things), that contribute to the disadvantaging of women. And no, this isn't a matter of pointing fingers.

    ReplyDelete
  13. The Scientist article was so unscientific it is incredible anyone took it seriously. The opening premise was based on finding: "only about a fifth -- 22% -- of the bloggers at scienceblogs.com are women." First falsehood: believing scienceblogs.com is about science. Bloggers selected by a media company for the viewpoints they espouse and being paid to write them doesn't make them scientific.

    That said, women bloggers, even those with far more credentials and years of experience to write on the issues they discuss, are shut out of the science and medical blogsphere boys clubs. Male bloggers seem to dominate the field publicly, but the public may not see them as much the words of real science as they believe themselves to be.

    ReplyDelete
  14. You said:

    Do you really think that in 2008 there are significant numbers of male scientists who are that biased against women? I don't.

    I shouldn't have to say this, Melissa, but I am not denying that sexism exists.

    Larry, so are you saying that sexism exists but it's not significant? Because I disagree with that completely. I disagree with your OP based on not only my own personal experiences, but on research of gender in the field and anecdotal stories from other female scientists.

    As far as I can tell you disagree with the study based on the fact that it just "doesn't sound right to you." Do you recognize that as someone who would never have been on the receiving end of sexism that you may have an observational bias here?

    You seem to be overly defensive and are inferring things from my post that aren't there. I never wrote that all men are sexist, not even most men. Yes, D is an exceptional case, but not as exceptional as you may want to believe.

    How will men and women work together to combat gender bias if we don't recognize the initial fact that gender bias does exist, and at a significant rate to impact the average female in the field? It only takes one bigoted person to derail a student's career path.

    ReplyDelete
  15. How in the world are scientists ever going to have a serious discussion about this issue if we begin with the assumption that one side is right and anyone who disagrees is sexist? Is that how real scientists are supposed to behave?

    It's not an assumption. The folks charging sexism in peer review have data. Insufficient data, of course. On the other hand, you have nothing beyond anecdote-based gut feeling:

    This does not sound right to me. There may be all kinds of reasons why women don't publish as much as men (if it's true) but I doubt very much that reviewers are more likely to reject a paper by a women scientist.

    You doubt it. Very good. I doubt that you've thought about the question with any care at all. That you don't point to the relevant literature as you would on a question of biochemistry underscores the laziness of your assertion (it doesn't rise to the level of argument).

    George, please answer a question. Do you believe everything you read in the scientific literature?

    Of course I don't. But I do think that some data is a better starting point than no data. You might do well to ponder the difference between skepticism and denialism.

    ReplyDelete
  16. Larry, conscious bias is not the main problem, at least among scientists. But we all (men and women) have unconscious biases too, and these strongly disadvantage women.

    Virginia Valian's book Why So Slow has a lot of the data.

    ReplyDelete
  17. This sounds like a problem with a clear solution: an overhaul of the peer review system.

    We need either absolute secrecy (i.e. double-blind) or absolute non-secrecy (i.e. identify the reviewers). The current system is broken, end of story.

    ReplyDelete
  18. While double-blind reviewing sounds nice, it isn't practical. Because of self-referential tendencies, not to mention the trend to write papers in active rather than passive voice, a reviewer could identify the lab(s) the manuscript came from the great majority of the time.

    ReplyDelete
  19. melissa says,

    Larry, so are you saying that sexism exists but it's not significant? Because I disagree with that completely.

    I suppose it depends on what you mean by "significant." Do I think that the vast majority of male scientists are sexist? No, I do not. Do I think that a simple majority are sexist? No, not in my (admittedly biased) experience.

    I disagree with your OP based on not only my own personal experiences, but on research of gender in the field and anecdotal stories from other female scientists.

    Funny, but those are exactly the same things that I use to form an opinion—except that I'm adding in the stories from male scientists as well.

    I'm not discounting the personal experiences of many women in science. Those experiences are significant and real. There's a problem that needs to be addressed by focusing on those (male) scientists who are at fault.

    As far as I can tell you disagree with the study based on the fact that it just "doesn't sound right to you." Do you recognize that as someone who would never have been on the receiving end of sexism that you may have an observational bias here?

    I recognize that lots of people have observational biases. It's part of human nature.

    Melisssa, it almost sounds as if you don't want me to discuss this issue because, as a male scientist, I have never experienced the problem first hand. Is that what you're saying?

    How will men and women work together to combat gender bias if we don't recognize the initial fact that gender bias does exist, and at a significant rate to impact the average female in the field? It only takes one bigoted person to derail a student's career path.

    The way we can work together to solve the problem is for women to make the distinction between widespread systemic sexism, which accuses all men—and women, for that matter—and problem individuals, who we can eliminate.

    Yes, one bigoted man can derail a student's career path and that's unacceptable. On the other hand, one good man can make all the difference. Please don't lump the good men in with the bad apples, even if it's only by implication.

    You don't want all women to be treated as inferiors and I don't want all men to be treated as sexist slobs. It seems to me there's opportunity to work together as long as we both agree to those conditions.

    ReplyDelete
  20. rosie redfield says,

    Larry, conscious bias is not the main problem, at least among scientists. But we all (men and women) have unconscious biases too, and these strongly disadvantage women.

    Do all men have unconscious biases that discriminate against women?

    If so, how do you suggest we can deal with those "unconscious" biases?

    Do you have unconscious biases against male graduate students and male post-docs in your lab?

    ReplyDelete
  21. Funny, but those are exactly the same things that I use to form an opinion

    Then you should post references to research and evidence which supports your claim. There is a wealth of research on the subject of gender biases in the sciences, and it's clear from your outright dismissal of GrrlScientist's claims that you are not familiar the subject.

    Melisssa, it almost sounds as if you don't want me to discuss this issue because, as a male scientist, I have never experienced the problem first hand. Is that what you're saying?

    No, not at all. I, for example, have never experienced racism in my field. From my experience, and the experience of all my white friends, racism doesn't exist at a significant rate for it to affect minorities...

    But of course, I have no clue what I'm talking about. I have no experience with racism because there's no reason for someone to be racist against me. For me to form an informed opinion on the subject, I would need to do research - read about the experiences of minorites and any trends in the field that point to an unconscious bias among people.

    On the opposite side, as a female, I worry that I may be biased, and I am careful to not point fingers without sufficient proof, or solely use my own experiences as a basis.

    I am not lumping all men into the same category. However, unconscious sexism happens with both men and women, and is a very real phenomenon.

    If you want to debate, then post evidence to back up your opinion, otherwise this is all meaningless. I'm done.

    ReplyDelete
  22. Do all men have unconscious biases that discriminate against women?

    If so, how do you suggest we can deal with those "unconscious" biases?

    ***********************

    One of the articles I cited above does delve into this. More follow-up research is necessary. One way to deal with biases though maybe to allow more time to think before making a decision. Rational thought to overcome biases isn't exactly an idea I think you would be against.

    Martell article in the Journal of Applied Social Pyschology (1991) (doi:10.1111/j.1559-1816.1991.tb00515.x)

    ReplyDelete
  23. While double-blind reviewing sounds nice, it isn't practical. Because of self-referential tendencies, not to mention the trend to write papers in active rather than passive voice, a reviewer could identify the lab(s) the manuscript came from the great majority of the time.

    I am as guilty as the next person of writing "We and others have shown..." or something to that effect in my manuscripts, so you're quite right: reviewers should be able to figure out who I am, even with DB reviewing.

    But right now there is precisely zero accountability on me as a reviewer, and I have no recourse against an obviously unfair review (which, believe me, I've had).

    Which is why I agree with Larry, that completely open peer review is the way to go.

    ReplyDelete
  24. Without reading all of the comments to this post, and without reading the article cited by Grrlscientist (since I don't have access to it) - it seems to me that the question of gender (or any other bias) is complicated by the large number of variables involved in peer review decision. Looking at only first author isn't likely to be indicative. I've been a professional scientist for 20 years, and served as peer reviewer (and author) on many papers. In some fields it is common practice for the professor to be the LAST author, and the first author position given to the student that needs to fill out their publication list prior to graduating. Most reviewers know this, so don't really pay attention to the first author. If they know the professor, they'll recommend the paper. I've witnessed this first hand in both the review of papers and grant applications. many new assistant professors complain that their established colleagues care more about the prestige of an author's university or research advisor that about the quality or merit of the work. These variable should have been controlled for in the research but don't seem to have been.

    Too often in grant reviews have I seen poor proposals funded because the author was buddies with the reviewers; and I've seen worthy proposals trashed because the project threatened the work of a reviewer or one of their pals. I've never seen a paper or grant rejected because of the sex of any author.

    From my personal experience, I have only seen sexism against males - in industry hiring. When I graduated with my PhD in Chemistry in 1996, it was the rule that the men would apply to dozens of positions before getting an interview, while the women would receive offers for every resume they sent out. In one case, a male and a female PhD from the same group applied to the same position, both had on-campus interviews with the same person from the company. The male candidate received a letter stating "there were many more applicants than positions available" so we won't hire you. The female received a letter from the same person stating "there weren't enough applicants to fill the number of positions available so please accept our offer." Many of the women scientists teased the men about this, too. Most of the men had to take post-docs while they continued to look for a professional position.

    As a final observation, I have yet to hear a satisfactory response to the question "What does it matter what the gender of the scientist is to the scientific result?" Are the laws of physics different for men versus women? Are superconductors only superconducting for men, not women? As a consumer of scientific progress, I am only concerned with the quality of the work and the product of the research - not the identity or the gender of the scientist that performed the experiment. Why has society forgotten this fundamental rule of science?

    ReplyDelete
  25. melissa says,

    Then you should post references to research and evidence which supports your claim. There is a wealth of research on the subject of gender biases in the sciences, and it's clear from your outright dismissal of GrrlScientist's claims that you are not familiar the subject.

    I started this thread so we could talk about the "research." In particular, we can talk about the paper that GrrlScientist selected.

    Yes, there are many other similar papers but this one seems like a good place to start.

    The authors show that the shift to double-blind reviews in one journal is correlated with an 8% increase in the number of female first authors. They suggest that this reflects an 8% bias against women when their identity is known; a bias that disappears when their identity is concealed.

    The authors suggest other possible explanations but it's clear that this is the one they prefer.

    The implications are obvious. Some reviewers apparently discriminate against female scientists and reject papers for publication on the grounds that the work was done by women.

    Incidentally, one other journal (Biological Conservation) showed a similar increase in the number of female first authors without double blind reviews. Here's how the article explains that observation.

    This variation might reflect a demographic shift in discipline of conservation biology, which might be independent of changes in behavioural ecology. It might also reflect variation in author submission behaviour, given the perception of increased anonymity. However although the most intuitive explanation is that the review policy reduced the potential for bias in the review process, double-blind review is not practised at BC, and increased submissions by females probably explains the observed pattern (R. Marrs, personal communication).

    In case anyone is interested, here's the raw data for each journal. The first number is the percentage of papers with women as first author in the period from 1997-2000. The second number represents the percentage from 2002-2005. The last number is the percentage change in female first authors.

    BE   27.6 36.9 +9.3%
    BES 32.8 33.6 <1.0%
    AB   34.3 39.5 +5.2%
    BC   20.9 28.2 +7.3%
    JB   22.3 20.5 -0.8%
    LE   23.8 31.2 +7.3%

    I'm on record as expressing skepticism about a number of published scientific papers. Just because it's in a science journal does not mean that it's correct.

    Why should this paper be immune to critical evaluation? The conclusion conflicts with everything I know about the review process and the behavior of my fellow scientists. It just doesn't make sense to me that large numbers of scientists would reject a paper just because one or more of the authors is a woman. We're talking about scientists who work in biology departments where the majority of graduate students are women. Chances are the majority of their own papers have at least one female author.

    This work is no different than any other study in the scientific literature. We need to be skeptical of everything that's published and we need to be especially skeptical of work that conforms to our pre-conceived bias since those are the very papers that can lead us astray.

    Let me add the appropriate disclaimer that seems to be required in these conversations. I'm not denying that some scientists are sexist. I'm questioning whether this is a systemic problem that applies to all male scientists. I'm questioning whether sexism is evident at the level of reviewing papers for publication.

    I can tell you for a fact that I don't care about the gender of authors on the papers I review. As a matter of fact, I don't even pay attention to their gender. I had to back and look at the papers that I reviewed last year to see if there was any evidence of a correlation. (There wasn't.)

    ReplyDelete
  26. . I'm questioning whether this is a systemic problem that applies to all male scientists.

    So we go from an argument from personal incredulity (a consistently taken approach in your posts on this thread) to the triumphant defeat of a strawman.

    At this point there is more than anecdotal evidence that you are trolling, Larry.

    ReplyDelete
  27. Sexist bias isn't always the stereotypical bigot or overt bigot seeing a lady's name and being fully aware of how little they think of women in science. It can be far more subtle, and women can do it, too - masculinity is valued in appearance in the office place, women wearing masculine clothing tend to be better-recognized and promoted faster. This type of bias could also be happening in the world of science, to a lesser extent - Sheryl Bergman (fake name) wrote a paper for a specific type of genetic analysis? If the person registers a sex with the name, bias immediately enters in some form or another whether we like it or not. Past experiences could inform you to treat the paper especially carefully, as it's a woman's science paper and deserves your greater attention, leading to greater scrutiny, or many other such things. This analysis should be confirmed at other journals, but the methodology seems interesting and elegant - eliminate bias simply in name identification and see how the results change.

    ReplyDelete
  28. g says, "I'm questioning whether this is a systemic problem that applies to all male scientists.

    So we go from an argument from personal incredulity (a consistently taken approach in your posts on this thread) to the triumphant defeat of a strawman. "


    What strawman? GrrlScientist, the author of the blog article being, comes to the conclusion that: "Basically, science is still a very sexist community where its female practitioners publish less frequently than men at least partially because of the peer-review system that is in place. I think the commonly used single-blind peer review process is biased against papers whose lead (or sole) author is female, just as the field of science is biased against women in general."

    OK, so replace Larry's "all" with "most", since we're talking about general bias in science against women. The skepticism still seems to hold. Or are you arguing that it's not just men but other women who are also biased against women in science?

    ReplyDelete
  29. G says,

    So we go from an argument from personal incredulity (a consistently taken approach in your posts on this thread) to the triumphant defeat of a strawman.

    Could you explain this comment?

    The argument from personal incredulity is not the same as legitimate skepticism. I presume you have looked at the study in question. Do you honestly think the data points to systemic bias against female first authors on the part of reviewers?

    I strongly suspect that some of the believers in that study are guilty of a little too much credulity—otherwise known as bias.

    Your statement is a response to my objection that "this is a systemic problem that applies to all male scientists." Rather than just toss out the word "strawman" why don't you tell me whether you agree or not?

    According to you, what percentage of male scientists are biased against women?

    ReplyDelete
  30. Shirakawasuna says,

    This analysis should be confirmed at other journals, but the methodology seems interesting and elegant - eliminate bias simply in name identification and see how the results change.

    If your analysis is correct and there is widespread bias against women in science, then we should be dealing with the problem instead of devising tricks to get around the symptoms.

    A double-blind review process does nothing to eliminate bias against women. All it does is hide it in one particular manifestation.

    I remain unconvinced that there's significant bias against women during the review process. But if there is, then we should identify the relevant reviewers and remove them from the process. How hard is that?

    ReplyDelete
  31. I just looked at the data and they didn't measure the right thing in order to test their hypothesis.

    They measured the percentage of papers published by females. What they should have measured was the percentage of female authored papers rejected by single and then double-blind review. If double-blind review resulted in statistically significant less rejection then would provide fairly strong evidence for a bias against women (although alternate explanations are still possible).

    If you really wanted to be thorough you could then measure the relative rejection rate of male and female reviewers on male vs female authored papers.

    ReplyDelete
  32. I agree with you Ryan. I think that's a serious methodological flaw in the paper. Now that I've had a chance to look at it more closely and read the comments here and on GrrlScientist's blog, I'm beginning to wonder how this paper ever got past the reviewers at Trends in Ecology & Evolution.

    ReplyDelete
  33. Larry Moran says:
    "If your analysis is correct and there is widespread bias against women in science, then we should be dealing with the problem instead of devising tricks to get around the symptoms. "

    I agree. I only meant that the study itself was very simple and elegant to perform, so confirmation/refutation should be easy to get.

    I too am not convinced that bias is that strong nor ubiquitous, but there are biasing factors we often don't think of immediately and I only listed a couple in my previous comment. At the same time, this study doesn't seem terribly dubious other than doubting the results based on earlier experiences and the small and specific sample size.

    As always with studies into social behavior, more research is needed...

    ReplyDelete
  34. Hmm, after reading Ryan's comments, they did indeed neglect that control. Darn statistics.

    ReplyDelete