Another study shows gender bias in science

[rowan]

Another study shows that with exactly equal qualifications, woman are rated as less competent, less hirable, and less likely to be mentored (after hire).

To test scientist’s reactions to men and women with precisely equal qualifications, the researchers did a randomized double-blind study in which academic scientists were given application materials from a student applying for a lab manager position. The substance of the applications were all identical, but sometimes a male name was attached, and sometimes a female name.

Both men and women ranked the women applicants lower. This shows how deeply ingrained our cultural biases are. The author of the Discover article says at the end "I have no reason to think that scientists are more sexist than people in other professions in the US, but this is my profession, and I’d like to see it do better" but I suspect if we actually probed that more deeply we would find that science is slightly worse than other professions, simply for the fact that it won't admit it has a problem. I would like to see that kind of study done, though it would be more complicated to do.

This is along the lines of many other things I've quoted/mentioned before, such as peer-reviewed articles, and hiring decisions at higher levels (e.g. for tenure track positions). Unsurprising, for those of us who follow the literature, but another (depressing) data point.

http://blogs.discovermagazine.com/cosmi ... =pulsenews

Links to original study (and others) in the article.

[Sonic#]

I think these studies are really good (not to mention necessary) because it's so easy to say that we would treat people the same despite a difference in gender, race, age, or appearance, but so hard to actually practice that. It's hard to definitely demonstrate that a form of discrimination that is primarily perceptual is going on, and thus hard to even have a conversation about the problem. I hope this is a piece that continues to work at the gender biases in science.

One commenter suggests a double-blind selection process as one way to combat the problem: all identifying information would be removed, leaving only one's credentials. That would work for part of the selection process, though it's impossible to do that for interviews.

Finally, I did a search on Google Scholar on a whim, and came up with this study, which looked at gender bias in citation for dendrochronologists (those people who use trees as time machines): http://www.bioone.org/doi/abs/10.3959/2009-10.1. The surprise? There's no gender bias in citations. According to the abstract, a couple of reasons for the equality of citation include frequent co-authorship between authors of different genders as well as the productivity of women in dendrochronology. In the fields you work in, rowan, does that explanation seem to work or apply to other fields

[lillerina]

I suspect that people who work in fields which pride themselves on rationality would be less likely to admit to potential bias because people who like to think of themselves as strictly rational beings don't like to think that this sort of messy feelings bullshit could apply to them. This may also link to the relative lack of women in STEM fields.

I've seen studies that suggest not only that men's competence is rated more highly than women's with equal qualifications but that modesty is rewarded in women and not in men - actually a man who displays modesty is seen as weak, in much the same way that a woman who does not display modesty about her accomplishments is seen as aggressive, pushy and bitchy.

[rowan]

One commenter suggests a double-blind selection process as one way to combat the problem: all identifying information would be removed, leaving only one's credentials.

I think that it would be good to implement this as far as we could. However, even with double-blind selection process in place, once you get down to specialized sub-fields it's pretty impossible not to know who it is. However it would be helpful for less specialized things.

don't like to think that this sort of messy feelings bullshit could apply to them

Very much a prevalent attitude, from what I've experienced. (And I will note that at one point I thought so too!)

actually a man who displays modesty is seen as weak, in much the same way that a woman who does not display modesty about her accomplishments is seen as aggressive, pushy and bitchy

*nod* I'm not sure how this plays out in physics, exactly. I've seen this in larger population studies. It would be interesting to see how that carries into our rarified atmosphere here... where 'interesting' probably means 'depressing'

Sonic, I'll have to look at your links a little later... (so you know I'm not ignoring that question)

ETA: Ok, I've looked and it's tickling in the back of my head that there was some research on that kind of thing... but I don't remember exactly what...might have been something about how many co-authors etc? I'll have to poke around. I do not have a good feel for how things are cited in my field but I would expect that women on average tend to have more co-authors than men do. However that is just a gut feeling, and also since we're SO outnumbered (11%) that might be somewhat meaningless without additional information (for example I know there is a trend in general towards more collaboration - so old dudes will have more single-author papers, and then you'd have to take into account some normalization for age/years in the field or something).

[Aum]

It's the whole boys club mentality. Men are brought up to think that men are from mars and women from Venus, so they'd rather work with their "kin" because they'll relate better and get more done.

Plus science has recent history of being incredibly sexist, even in the way it describes scientific phenomena.

[monk]

[quote="Sonic#"]: http://www.bioone.org/doi/abs/10.3959/2009-10.1

Notice when you look at the references cited in Sonic's link that the author names are first name initial and last name only. Omitting the full first names from citations (which is pretty common) likely goes a long way in removing gender bias.

[androkguz]

Guys, when you criticize this bias, are you implying which of the following four options?
1) There is no correlation, not even statistical, between being male/female and being hirable/mentor-able/competent for a job.
2) Scientists have this correlation wrong.
3) The correlation exists, but it is wrong to act on it. Or in this case, just point it out.
4) Scientists, both men and female, would just rather not work with women if possible.

[Sonic#]

I believe 1, but whether or not 1, one's gender should not interfere with assessing one's qualifications. If the same qualifications are judged differently because the gender of the applicant is different, then there's some bias that interferes with making the best choice.

Omitting the full first names from citations (which is pretty common) likely goes a long way in removing gender bias.

I hadn't thought of it that way. MLA citation uses last names and then first names. My adviser knows many of the same scholars I do, but having their names may have prompted my adviser to observe that I was tending to cite female scholars in my prospectus. Which is true, though I hadn't consciously noticed it. Anyway, I imagine that would have a greater effect for the reader of an article's citations than for its author though, since the author would presumably look at the full articles and then work from there.

[rowan]

Guys, when you criticize this bias, are you implying which of the following four options?
1) There is no correlation, not even statistical, between being male/female and being hirable/mentor-able/competent for a job.

If you read the article, they used identical resumes, and just changed the names. They asked all of the people evaluating the resumes to rank them in terms of hireability, competency, and whether or not they would be willing to mentor this student. Therefore in the study, there was an enforced parity between male and female competencies etc.

2) Scientists have this correlation wrong.

It's pretty obvious that it's a correct conclusion.

3) The correlation exists, but it is wrong to act on it. Or in this case, just point it out.

Why would it be wrong to point it out or act on it? It's clearly showing subconscious bias, and something we should strive to eliminate.

4) Scientists, both men and female, would just rather not work with women if possible.

It is likely not a conscious bias in most of them (though I have heard enough stupid things said to know it's certainly conscious for some scientists). I'm sure if you asked most of them, they would honestly say that they have no preference for working with either men or women.

Omitting the full first names from citations (which is pretty common) likely goes a long way in removing gender bias.

I agree, and it's why I've always used my initials when publishing (yes, this was a conscious decision based on my knowledge of gender issues). Whether or not it's common depends a lot on your field. In my field it's becoming much more common, so that doing so doesn't stand out, but there are still lots of people who use their full first name. (In part I think it's becoming more common because when you have a team of 40+ people you don't want to type all those first names out )

[Neko]

This is a little off topic but related to gender bias: I realized I had been programmed with that bias when I found out the author of a series I enjoyed was female--she used her first and middle initials on her books. I've read lots of female authors before but the fact that I was surprised a woman had written it was pretty telling. I thought the author was a man by the way it was written and, really, just by default.

I don't think anyone has anything to lose by examining their own bias. It's unfortunate that pride seems to get in the way of that but I'm optimistic as long as it continues being studied and discussed. As a woman with education and career plans in a STEM field, it won't always be easy to keep that optimism.

[androkguz]

@Sonic: It is interesting, your first sentence could be read "I believe 1, but if not 1, then 1".
I am pretty sure that option 1 is very unrealistic. There is a correlation between pretty much every aspect of one's life. I'm sure that if you made a study like this where the only difference was the type of music that the person preffers, people who like reaggeton would get less points than others.
The fact is you don't calculate a person's c/h/m by using an objective formula taken from concrete measurements. You do it by estimating how the person is and how he/she would behave given the information you have. If we lived in a world where 1/10 male scientists liked to solve problems with fists fights while only 1/600 female scientist did, then logically that would make "unknown female scientists with resume A" more comp/hir/ment. than "unknown male scientist with resume A". Sure, it would be even better to check the person personally and try not to discard it before hand, but that's really not posible in this investigation.

@rowan: I think you didn't understood my question. What this article concludes is that on average, people in the scientific community value being a guy 0.6 points higher in competence, ~0.8 points in hireability and ~0.6 points in mentor-ability than being a gal.
This is in the situation where no other personal information is given besides what goes on a resume: that is, academic stuff.

This bias implies that the community on average considers, conciously or not, that there is some correlation between being a man/woman and being competent/hireable/mentorable. For instance, I can image some people would think that being a woman is a somewhat less desirable characteristic for an employee because she could get pregnant and leave while still being paid.

What I was asking was the reason you critizice the bias.

[Mathmo]

This bias implies that the community on average considers, conciously or not, that there is some correlation between being a man/woman and being competent/hireable/mentorable. For instance, I can image some people would think that being a woman is a somewhat less desirable characteristic for an employee because she could get pregnant and leave while still being paid.

What I was asking was the reason you critizice the bias.

Because it is completely unjustified and does a lot of harm. (In particular, since this is often brought up, not all women want/have children and it is very unfair to not hire them because you think there's some chance that they might; also a man could become a father and leave while still being paid (depending a bit on country and paternity laws)

This is a little off topic but related to gender bias: I realized I had been programmed with that bias when I found out the author of a series I enjoyed was female--she used her first and middle initials on her books. I've read lots of female authors before but the fact that I was surprised a woman had written it was pretty telling. I thought the author was a man by the way it was written and, really, just by default.

That's kind of happened to me too - I was surprised that Malorie Blackman is female (I'd read several of her books - which are excellent, they're YA books which explore racism via a world much like ours but with many racial assumptions reversed - and then heard her on the radio). All the female spellings of the name I'd come across before had a 'y' in them (e.g. Mallory) so I'd assumed that 'Malorie' was a male spelling, even though that didn't quite make sense when I thought about it.

[Sonic#]

@Sonic: It is interesting, your first sentence could be read "I believe 1, but if not 1, then 1".

Then it could be read wrong. 1 is a statement of correlation. The second half of my statement addressed how we verify competence. My claim was that, even if a correlation between gender and competence existed, then the gap in competence should be evident in the application materials, the history of publication, and so on. Thus, it requires no recourse to one's gender.

Indeed, using gender to judge someone on that basis would be foolish, since, as the article explains, the bias inflects how more specific measures of competence are received. That's a bad thing because the things in one's resume, writing sample, references, and publication history are eminently more relevant to the actual performance of a job than one's gender. A woman would have to be demonstrably better than a man to be received equally, which is a decidedly ass-backward way of doing it - the fact that she has better qualifications should mean she is regarded better, just as the fact that she has equal qualifications should mean that she's regarded equally competent for the position.

That's why.