Where are all the women? The math said there would be more women!
October 21, 2015 12:08 PM   Subscribe

One mathematician’s formula suggests that all-male lineups don’t “just happen,” despite what conference organizers might claim. "...in any conference with over 10 speakers, say, it would be extremely rare to have no female speakers at all—less than 5 percent chance, depending on one’s assumption about the percentage of women in mathematics as a whole."
posted by sharp pointy objects (72 comments total) 36 users marked this as a favorite
 
The phrasing of this is really weird - it makes it sound like some sort of original or sophisticated new mathematical model, but really, this is basically standard high-school level math. (As Prof. Martin would undoubtably be the first to agree.)
posted by kickingtheground at 12:15 PM on October 21, 2015 [14 favorites]


I once called out a theater conference where, for their prized, mainstage play readings, the number of women represented had always been low, and had dwindled to none at all. A former employee of the conference responded tersely, pointing out that the plays go through a blind reading process, so therefore it was impossible that sexism was at play.

I likewise pointed out the statistical unlikelihood of this happening, and then said that we are left with two choices: Either women are inherently worse playwrights, or there is some unexamined sexism at play.

But what? I was asked.

Oh, I don't know. The fact that the two people who actually selected the mainstage plays were middle aged white men who tended to select plays that reflected their experiences and interests, and that women weren't necessarily writing that sort of play.

There is at least one women who participates in the final reading process now, and the number of female playwrights represented had increased dramatically. But, no, before I brought this up, people actually believed that somehow it made sense that year after year there were no women represented in this way.
posted by maxsparber at 12:17 PM on October 21, 2015 [87 favorites]


" In the technology sector, there’s an almost evangelical adherence to the religion of meritocracy, no matter how many studies come out proving that we all have unconscious biases"

/me nods so hard her head hurts.
posted by Annika Cicada at 12:18 PM on October 21, 2015 [34 favorites]


Does it also calculate the likelihood of "Women in ..." panels that don't actually have women on them?
posted by zombieflanders at 12:19 PM on October 21, 2015 [16 favorites]


Does it also calculate the likelihood of "Women in ..." panels that don't actually have women on them?

This is a really nice example of the difference between theory and practice because in theory the number is really fucking low and in practice it is asymptotically approaching one.
posted by Mrs. Pterodactyl at 12:21 PM on October 21, 2015 [5 favorites]


I got an interesting answer to an AskMe question last week which said that female academics tend to be pushed toward "student care" roles, while male academics are pushed toward research and funding roles. That seems like it would be one place where systematic bias would ultimately lead to results like this.
posted by clawsoon at 12:25 PM on October 21, 2015 [3 favorites]


He should really present that paper...
posted by Nanukthedog at 12:28 PM on October 21, 2015


Last year when the "Mammoth Book of Mindblowing SF" was published and JUST COINCIDENTALLY HAPPENED to contain no stories by women or POCs, someone did a similar mathematical analysis. I quote the relevant bit:
[L]et’s seriously lowball the range by rounding down, and say that, during the period from 1958 to 2006, women contributed just 25% of all professionally published SF works. Now, The Mammoth Book of Mindblowing SF contains twenty-one stories – five of which, crucially, were new works commissioned especially for inclusion the anthology. That leaves us with just sixteen stories potentially drawn from the period Wright is referencing. And if you do the maths on the basis of these numbers – namely, if you were to pick sixteen stories at random from a collection where 25% were written by women – then 99% of the time, you’d end up with at least one female-authored story. Which means that Ashley’s anthology would have been more diverse if he had, in fact, chosen his works at random; but of course, the point is, he didn’t. Not only did he commission five new stories exclusively from white male writers, but in digging through the entire history of SF, he somehow managed to miss even classic greats like Samuel R. Delany, Ursula K. Le Guin, Octavia Butler, Joanna Russ, Andre Norton and James Tiptree Jr, as well as modern award-winners like Aliette de Bodard, Catherynne M. Valente and Elizabeth Bear. So, no: the case against Mike Ashley – or, more specifically, against The Mammoth Book of Mindblowing SF – was anything but “arbitrary”.
posted by Amberlyza at 12:28 PM on October 21, 2015 [41 favorites]


I just thought it was sadly hilarious that a mathematician has to point out to other mathematicians what should have been damned obvious to them.

Also, I love going to listen to people speak at conferences, but the older I get, the less interested I am in sitting and listening to nothing but old, white dudes talk about how Subject X revolves around their worldview/classification system/contributions/etc...

(Also thank you for commenting...this was my first post...social anxiety meltdown averted...)
posted by sharp pointy objects at 12:28 PM on October 21, 2015 [38 favorites]


Does this analysis assume that men and women in the pool of participants are equally skilled at math?
posted by andoatnp at 12:30 PM on October 21, 2015


Is there a reason to assume that people who have achieved the same academic and professional level would not be equally skilled based on gender?
posted by maxsparber at 12:31 PM on October 21, 2015 [36 favorites]


it makes it sound like some sort of original or sophisticated new mathematical model, but really, this is basically standard high-school level math. (As Prof. Martin would undoubtably be the first to agree.)

Yes! The truth is undeniably, embarrassingly, strikingly simple.
posted by Dashy at 12:34 PM on October 21, 2015 [5 favorites]


Technically yes but it's a bad reason.
posted by poffin boffin at 12:34 PM on October 21, 2015 [5 favorites]


> Does this analysis assume that men and women in the pool of participants are equally skilled at math?

Is it your assumption that people chosen to speak on a panel are chosen purely and solely on their mathematical skills? That seems a poor assumption.
posted by rtha at 12:38 PM on October 21, 2015 [17 favorites]


I just co-organized a mathematics seminar last Saturday. Out of 8 speakers, one was a woman. (She was also one of the two most renowned speakers.) One out of the four people on the organizing committee was a woman. (She was also the chair of the committee.)

There certainly is no lack of excellent female mathematicians. That said, the example in the link, ICM plenary speakers, is maybe not the best choice. The pool of people who are considered good enough for that is very small - it's maybe the most prestigious speaking honor in math. It may very well be that in the absolute top echelon, the percentage of women is far less than 24% - I don't think this has anything to do with innate ability, but with social factors and the fact that women are generally less willing to put their work about any other aspects of their lives, which is frankly necessary to make it to the absolute top.
posted by tecg at 12:43 PM on October 21, 2015 [2 favorites]


I don't think this has anything to do with innate ability, but with social factors and the fact that women are generally less willing to put their work about any other aspects of their lives, which is frankly necessary to make it to the absolute top.

Well then, institutional sexism it is.
posted by zombieflanders at 12:52 PM on October 21, 2015 [68 favorites]


Does this analysis assume that men and women in the pool of participants are equally skilled at math?

It's your assumption that women are less skilled than men at math? Because intentional or not, that's the insinuation I'm seeing here.
posted by happyroach at 12:53 PM on October 21, 2015 [2 favorites]


That seems a poor assumption.

C'mon, it's just Occam's Razor at work. The whole deal is that the women who weren't asked to participate in these conferences are simply less skilled than the men. A postulation of structural bias wilts in the face of the fact that women are notoriously bad with numbers.

And now that that's out of the way, we can trot out the same reasoning for women's exclusion from politics, art, and music.

I don't think this has anything to do with innate ability, but with social factors and the fact that women are generally less willing to put their work about any other aspects of their lives, which is frankly necessary to make it to the absolute top.

Less willing? Or less able?
posted by divined by radio at 12:54 PM on October 21, 2015 [6 favorites]


I think this one is among the most relevant points:
When addressing an event organizer (or anyone) who on meritocratic grounds opposes paying attention to gender, the crucial step is to draw explicit attention to their underlying assumption: they are assuming that the current system is purely meritocratic in practice, and that efforts to introduce gender into the decision-making is necessarily an addition of unfairness. Helping someone learn by presenting them with the truth, after all, will never work if they already have a conflicting falsity in their minds.
People truly, truly don't want to believe they are being anything but 100% meritocratic. For instance:

There certainly is no lack of excellent female mathematicians. That said, the example in the link, ICM plenary speakers, is maybe not the best choice. The pool of people who are considered good enough for that is very small - it's maybe the most prestigious speaking honor in math. It may very well be that in the absolute top echelon, the percentage of women is far less than 24% - I don't think this has anything to do with innate ability, but with social factors and the fact that women are generally less willing to put their work about any other aspects of their lives, which is frankly necessary to make it to the absolute top.

The entirely more likely explanation, that you can't make it to the top speaking events if you're in a sexist field which is sexist all the way down, isn't addressed.
posted by jeather at 12:57 PM on October 21, 2015 [22 favorites]


The point about unconscious bias - that we're so used to seeing more men on panels then women, and so we don't really notice that the current sitation is ... improbable ... without sexism somewhere in the pipeline until it is pointed out to us - is a good one. I'd like to think that I am a pretty strident feminist, and a feminist scientist, and a woman who has good ideas and notices misogyny and speaks up about it and works for equality and the like, but yesterday I took a few Project Implicit bias tests. Strong association between Male and Career, Female and Family, Male and Science, Female and Liberal Arts.

So I'm implicitly biased against myself, which is frustrating, but I'm also implicitly biased to behave in a misogynistic way when it comes to interacting with my male and female colleagues, and - someday when I'm important enough to be organizing panels and conferences - I'll be implicitly biased to recruit men for these opportunities. Now that I'm aware that this is a bias I have inadvertently internalized, it's something I can actively work against, but most of us don't recognize our inadvertent biases and that makes working to counteract that bias seem over the top and potentially problematic (MISANDRY! REVERSE RACISM! They're not on the panels because they're just not good enough).
posted by ChuraChura at 1:06 PM on October 21, 2015 [19 favorites]


Introducing correlational structure to statistical samples that are claimed to by independent and identically distributed is not like adding drops of pee into the ocean: it's like pissing into 8 oz of water and expecting someone to drink it.

I highly suspect that if you arranged the random variables involved in the decision given to people to speak at a math conference into a network (Markov field or bayes net), I would highly suspect that whole network percolates sparsely, and is extremely big (|V| > 10^6). That is, you might look at two variables and determine that that don't influence each other just by looking at it, but how do you not know that there isn't some path by which they influence each other, with path length >=3? There are a goddamn lot of possible paths, and there are few variables that can be set up to be really independent from each other. It's no coin flip.

This is like Schutzenberger attacking the basis of natural selection based upon his assumption of IID evolution events, or like psychologists declaring Gaussian determinants of the mind of all humanity based on 30 college freshman. In physics, IID is common as muck, but in every field that involves agents in any way, IID disappears, and is never to be found.
posted by curuinor at 1:16 PM on October 21, 2015 [2 favorites]


(I don't have anything against women in math, I just despise IID assumptions getting to places where they should not be, or arguments based on IID thought appearing anywhere where there is correlational structure)
posted by curuinor at 1:21 PM on October 21, 2015


Then I expect you really hate the claim that this guy's calculation is disputing on its own terms.
posted by Zalzidrax at 1:23 PM on October 21, 2015


The entirely more likely explanation, that you can't make it to the top speaking events...

I feel like many people here don't really want to engage in an open discussion, but (ironically) just want their pre-conceived notions reinforced. The thing is, there are lots of women in the elite of mathematicians, and they can and do of course make it to the top speaking events. Some evidence suggests though that they are underrepresented there compared to their 24% share of all Ph.D.s

As for the reasons of that, if you read my statement closely, I did suggest "social" reasons, which does encompass some form of institutional and societal bias against women. The fact is that making it to the top in math (and other fields in science) requires an almost manic dedication to your work that leaves little time for other things. In my experience, women are less willing to do so (and I certainly am not, just one of the reasons why I'm far from the top of my profession). Granted, the way society is set up, it's easier and more acceptable for men to live that kind of dedication.

Would be interesting to get some perspective from women mathematicians here.
posted by tecg at 1:25 PM on October 21, 2015


If you're interested in what sort of active steps you can take to remediate this - and a reference you can point other people towards in your organization - Facebook's "Managing Unconscious Bias" training materials are freely available and very good.
posted by mhoye at 1:27 PM on October 21, 2015 [7 favorites]


The power of a statistical check like this is that it provides a simple sanity check on the results. You can use your usual selection process and then run the simple check to see if you have totally screwed up; if (when?) you demonstrate that you have, you go back and see what you have missed. It wouldn’t take very much to improve things significantly, even if truly proportionate representation is a goal that is unlikely ever to be achieved, even if it could be correctly defined or applied.
posted by Quinbus Flestrin at 1:29 PM on October 21, 2015 [2 favorites]


C'mon, it's just Occam's Razor at work. The whole deal is that the women who weren't asked to participate in these conferences are simply less skilled than the men. A postulation of structural bias wilts in the face of the fact that women are notoriously bad with numbers.

I'm always amused when people try to argue that both
a) there is no structural discrimination against women, and
b) that women are worse than men

Both. at once.
posted by entropone at 1:34 PM on October 21, 2015 [12 favorites]


It's your assumption that women are less skilled than men at math? Because intentional or not, that's the insinuation I'm seeing here.

Yeah, because of sexism. Is there any reason to think that the sexism that leads women to be dramatically underrepresented in the field of math wouldn't also negatively impact their opportunity to achieve the same level of skill as men (while their male peers are benefitting from that sexism)?

There is endless evidence about the hostility to women within STEM fields. It doesn't have anything to do with the innate ability of women at math to say that institutional sexism has led to an situation that advantages male mathematicians, and that the outcomes we see reflect this.
posted by andoatnp at 1:47 PM on October 21, 2015


Does this analysis assume that men and women in the pool of participants are equally skilled at math

the fact that women are generally less willing to put their work about any other aspects of their lives, which is frankly necessary to make it to the absolute top.

I'm the top expert in my field on a very small subject, ABC. I know this because I pretty much invented ABC; for 10 years I was the only one who published on ABC even though others cited it; and whenever anybody has questions or needs help on ABC I get the e-mail. It's been a very, very small contribution to my field, but an interesting one with some value, in particular because other people and I have been able to use ABC in our research on bigger subjects. At this point I've been working on it (among other things) for ~15 years.

In the last two years, as the result of the progress overall that has occurred since I first published on it, ABC has suddenly emerged as a possible important application in a very big research subject. Enough so that at our biggest conference this year they decided to have a lunchtime panel on ABC. Which, for that conference, was a super big deal.

It was a 5 person panel. The panel organizer was male. The person in charge of scheduling was male. And guess what? Every single person on the panel was male. About ~95% of the panel was my research.

I wasn't invited. I'm a woman.

There was no rational reason for me not to be invited. There was some protest about this prior to the conference by multiple people, who were all told that it would be embarrassing to dis-invite any of the panel participants. (And placed me in the horrible position of having to choose between saying something or not.) It turned out to be even more embarrassing, because during the Q & A session afterward, I ended up having to be asked to answer about 1/2 of the questions. In fact, after the 3rd such question, the questioner actually asked point blank why I wasn't on the panel as I had done all the research!

And here's a wonderful side effect: now instead of getting emails about ABC, I get FORWARDED emails from those 5 guys on the panel because they were positioned as the visual representation of ABC research.

So yeah, go ahead and tell me that it's because I'm equally unskilled. Or that I made a CHOICE. Go ahead, tell me that again. But before you do, I suggest you read and learn something about all the obstacles women face in STEM, many of which have been discussed here on MeFi. Read the bitch in business thread I linked above. Read the emotional labor thread about the pressures that women face when it comes to taking care of other people.Then come back and tell me it's because I lack skills or that I make choices that keep me from succeeding.
posted by barchan at 1:48 PM on October 21, 2015 [284 favorites]


entrope, DVR was being sarcastic

Now I have to go take a walk.
posted by barchan at 1:50 PM on October 21, 2015 [1 favorite]


Well curuinor, you're probably correct (if my female brain understands your complaint) that there is a network effect on women in science. Something maybe like women are cited less, either by themselves (pdf) or others (study pdf)? Then that translated to them not doing as well in science as men and dropping out. So when the old boys network gets activated to get 'high quality' women to speak at a mathematics conferences, there are hardly any?

All models are wrong, some are useful.

You could definitely make a model that was complex enough to explain exactly why women aren't being selected to speak but why would you? There are plenty of reasons, we know a lot of them already. We don't actually want a great model (for modelling's sake), we want to fix the problem.
posted by hydrobatidae at 1:51 PM on October 21, 2015 [5 favorites]


I feel like many people here don't really want to engage in an open discussion, but (ironically) just want their pre-conceived notions reinforced. The thing is, there are lots of women in the elite of mathematicians, and they can and do of course make it to the top speaking events. Some evidence suggests though that they are underrepresented there compared to their 24% share of all Ph.D.s

. . . because of sexism.
posted by jeather at 1:55 PM on October 21, 2015 [1 favorite]


So I'm implicitly biased against myself, which is frustrating, but I'm also implicitly biased to behave in a misogynistic way when it comes to interacting with my male and female colleagues, and - someday when I'm important enough to be organizing panels and conferences - I'll be implicitly biased to recruit men for these opportunities.

Isn't it wild, how it feels so irretrievably tentacled inside your brain? This is why I especially hate it whenever dudes use evidence of this kind of subconscious bias (internalized misogyny) at work to be like, "See? THIS woman doesn't have a problem with [manifestation of sexism], CHECKMATE FEMNSINTSS."

Setting aside the vast individual profits that can be acquired via conscious patriarchal bargaining, women aren't just subconsciously socialized to mistrust and dislike ourselves and each other, we're very explicitly rewarded for it. Women who agree with men are the only kind of women who are universally taken seriously, at face value. When a woman's stance aligns with a man's, especially when it supports a male supremacist point of view, that's pretty much the only time you can be sure she's going to be believed. And damn, it feels really good to believed! So it turns out to be a self-perpetuating cycle. Women are lauded for keeping other women down, including ourselves, and dissuaded from complimenting or standing up for ourselves because that's just bitchy, shrill, self-centered, pearl-clutching, acting too big for our britches, etc. So it can easily progress to the point that, because it can be so beneficial for individual women in any male-majority environment to talk shit on other women in the same scene, we start to associate the denigration of women with personal or professional reward even as we understand it as engagement in our own continued subjugation.

Along the same lines, that's why "you're just one of the guys" isn't remotely value-neutral; it's only ever intended as one of the highest forms of compliment, in the same way that "you [perform any action or behavior... like, I dunno, do math] like a girl" is only ever intended as a grave insult. Also why "I'm not like the other girls/women" is actually just a way of reifying your respectable dissimilarity from an otherwise indiscernible sea of despicably female nitwits, and done very pointedly to curry favor with misogynists who hold positions of power and influence. Theoretically, it's supposed to be some kind of assertion of individuality, but it doesn't perform that way: The world is surprisingly full of women who are very quick to insist that they aren't anything like any of the others.
posted by divined by radio at 2:01 PM on October 21, 2015 [57 favorites]


The network is on the random variables, not on people. So induce a causal network on the causation of how women are underrepresented - sexism is a node (or a complex of nodes), it has a causal relation (or many, many causal relations) towards fewer citations, citation # is a node, and citation # is the one that has the causal relation towards being on panels. Now, add a crapton of nodes and a crapton of causal relations (edges).

If you are assuming independence, you are assuming that there are no causal relations (correlation is not necessarily causation, but discorrelation is necessarily discausation)

The thing that the easy coinflip model throws out, then, is the whole ecosystem of causation that keeps the imbalance in place, any possible Markov blankets that you can actually claim conditional independence if you know, the improbabilities that we want to keep.

And you do want to keep improbabilities: ceteris paribus, a randomly chosen person knows nothing about even something pretty widely known, like rings: you would have a Bernoulli variable with heads = 0.001 for getting a randomly chosen person out of the whole pool, so having 5 people together who all know what rings are is declared an impossibility. Obviously, the fact that these people all know their fields would be declared an impossibility there.
posted by curuinor at 2:04 PM on October 21, 2015


I will not type a comment simply consisting of "AAAAA" even thought that is how I feel about this today.

So, I'm not a mathematician, but I'm a theoretical physicist which has some similarities. I am on an organizing committee for an upcoming conference, and we're trying to come up with names to invite. Two of the people on the committee produced 10-person lists with no women.
For me, I know that institutional bias is real, so I did my best to fight it, trying to make sure I got as complete a list to start from as possible, and then I culled to 10 names. I did not drop any women in the culling. I still only had 2 women on my list, and I was really trying. I know it's hard, but I have to say-- I kinda judge the hell out of the fellow committee members who clearly *didn't even notice* that their lists contained only men.

It's frustrating, even conscious effort is insufficient-- and I don't want to have to fight it all the time, partially because I want my own career instead of being shunted into the academic caring roles mentioned above.

But even more frustrating is it's clear there's some people who aren't even trying (damn, barchan, hope that committee got some blowback for their mistake! wtf!).
posted by nat at 2:12 PM on October 21, 2015 [6 favorites]


[Comment removed; long-form "I've been drinking and here's my thoughts on sexism" is a really bad direction to go with this.]
posted by cortex (staff) at 2:16 PM on October 21, 2015 [36 favorites]


Since the main point of the article was that there isn't equal representation from both genders on conference panels, I honestly don't think (for many, not all) conferences, that all the speakers have to be at the very top of their fields. In fact, I would bet many of the men who are invited now aren't. Barchan's experience is pretty much the perfect example.

So, why shouldn't there be some sort of speaker/panelist "affirmative action"? There are thousands of conferences, so thousands of speaker slots. You literally can't have only the top of each respective field as speakers, because there's only so many days in a year, and I'm sure they want to actually get some work done.

So offer X-amount of these slots to women, make it normal for there to be women in these spaces, even if it seems odd at first. I bet that many times they'll have plenty of experience and insight to offer, and if they sometimes don't, who cares. We're literally fighting to simply normalize the experience of seeing women in science, making them visible. It's been the same old tired status-quo for so long that at this point I'll take whatever I can get.
posted by sharp pointy objects at 2:19 PM on October 21, 2015 [5 favorites]


Introducing correlational structure to statistical samples that are claimed to by independent and identically distributed is not like adding drops of pee into the ocean: it's like pissing into 8 oz of water and expecting someone to drink it.

Well his point is that they are not, isn't it? You could argue that this is a bit of a straw man I guess.
posted by atoxyl at 2:23 PM on October 21, 2015 [1 favorite]


I honestly believe that "Women choose not to devote the time to be the top of their fields" is the modern sexism equivalent of "Really, negros love to work on the plantations."
posted by happyroach at 2:25 PM on October 21, 2015 [38 favorites]


Yeah curuinor, I think I get what you're saying - you think that a 'coin-flip' statistical models is not correct for this situation because of confounding effects on how a speaker is chosen.

What I'm saying is that this simple model is useful for pointing out the first step: that the end result is messed up.

We're still at the 'getting others to acknowledge there is a problem' stage.
posted by hydrobatidae at 2:28 PM on October 21, 2015 [7 favorites]


We're still at the 'getting others to acknowledge there is a problem' stage.

And when there's people even in this thread going "Oh no, it's just women don't WANT to be top mathematicians", you can see just how goddamn far we have to go.
posted by happyroach at 2:37 PM on October 21, 2015 [24 favorites]


barchan's comment just caused me to have a brain aneurism. If we ever meet, barchan, I'm buying you drinks. Strong ones.

I spent the majority of last week at the Grace Hopper conference, listening to many MANY stories just like this, from a ridiculous number of brilliant and talented women. It's not our choices, it's a structural issue. There are plenty of studies showing this, but if you're interested in your own internal biases, take a few of Harvard's tests at Project Implicit. (you can enter as a guest on the lower right).
posted by blurker at 2:39 PM on October 21, 2015 [10 favorites]


Yeah curuinor, I think I get what you're saying - you think that a 'coin-flip' statistical models is not correct for this situation because of confounding effects on how a speaker is chosen.

I think his point is just that this is a bit of an abuse of statistics for rhetorical purposes - Martin is knocking down a hypothesis that isn't necessarily something that people propose to be true. The counterpoint would be I suppose that it does illustrate in broad strokes just how weird the pattern of speaker selection is.
posted by atoxyl at 2:50 PM on October 21, 2015


What I just said is pretty similar to what you already said, sorry.
posted by atoxyl at 2:53 PM on October 21, 2015 [1 favorite]


Women choose not to devote the time to be the top of their fields

Women choose not to devote the time to all the seventeen layers of bullshit hell that a woman has to go through from primary school through to her PhD just to be able to reach equal ground with men who also want to be the top of their fields.
posted by Thella at 3:35 PM on October 21, 2015 [26 favorites]


I run an annual conference on a very specific subject in a male-dominated field. For the last 5 years+ I've made a conscious effort to have more women on panels, and to specifically remind every woman I know in the field that they should attend. The number of female attendees has been steadily rising, which I'm happy to see. And yet this year I couldn't get a single female panelist, despite making every effort. All of the 12 women I invited to speak (total speakers ~ 25) ended up declining. Most not because they didn't want to come, but because their companies or organizations weren't willing to let them, or weren't willing to pay to send them. It's incredibly frustrating as an organizer too...
posted by gemmy at 3:45 PM on October 21, 2015 [2 favorites]


There are already women who are producing work of equal and greater quality to the men at the top of their fields. They are getting recognized, winning grants, being hired, and publishing papers at lower rates than their make colleagues because of sexism, not because they just aren't dedicated enough! They (we) are producing this work despite the overt and covert sexual harassment they experience, despite these deficiencies in opportunities, and in spite of all the folks saying "well, they just don't have the same drive."
posted by ChuraChura at 4:00 PM on October 21, 2015 [21 favorites]


Oh barchan, that you are not wearing a prison jumpsuit while cleaning the jugulars of those men from your teeth is truly a testament to your strength. Go you.
posted by SecretAgentSockpuppet at 5:15 PM on October 21, 2015 [9 favorites]


Just an aside - My understanding from a biographical article about Grace Hopper was that she tended to downplay structural aspects of gender inequality, and favored individualist means as the solution. That point in the article left a strong enough impression on me that I have thus far stopped citing her as a role model for diversity in STEM. It's just a curious detail—that I've not been able to reconcile with more information. Maybe Hopper's biographies will shed light on this.
posted by polymodus at 5:18 PM on October 21, 2015 [1 favorite]


When Anne Churchland*, computational neuroscientist, was faced with putting together panels for the annual conference on Computational and Systems Neuroscience (or "CoSyNe," for short), she began compiling a list of women scientists in our area. It's called Anne's List, and she has made it available online so that other conference organizers can use this handy list.

*Daughter of famous philosophers Patricia and Paul Churchland, but a respectable thinker in her own right.
posted by tickingclock at 5:38 PM on October 21, 2015 [4 favorites]


I honestly believe that "Women choose not to devote the time to be the top of their fields" is the modern sexism equivalent of "Really, negros love to work on the plantations."

I totally, totally agree with this. And it isn't as though there haven't been a shitload of studies on this, showing that it has not really much at all to do with "choice" for a number of reasons. Here's a great one on HBS MBA graduates and the career gender gap:

Rethink What You "Know" About High Achieving Women (various points from the article, any emphasis mine):

Despite the fact that men and women actually have pretty similar career priorities, the belief that women value career less is widespread. We found that 77% of HBS graduates overall—73% of men and 85% of women—believe that “prioritizing family over work” is the number one barrier to women’s career advancement...


Our survey data and other research suggest that when high-achieving, highly educated professional women leave their jobs after becoming mothers, only a small number do so because they prefer to devote themselves exclusively to motherhood; the vast majority leave reluctantly and as a last resort, because they find themselves in unfulfilling roles with dim prospects for advancement. The message that they are no longer considered “players” is communicated in various, sometimes subtle ways: They may have been stigmatized for taking advantage of flex options or reduced schedules, passed over for high-profile assignments, or removed from projects they once led. One alumna, now in her late fifties, recalled, “I left my first job after being ‘mommy-tracked’ when I came back from maternity leave...


Time out of the workforce
could account for the fact that women are less likely to be in senior positions. After all, it’s often argued that because being in senior leadership is directly tied to years of professional experience, women are less likely to be in those roles precisely because they are more likely to have taken such breaks. So we delved deeper, with controls for variables such as age, industry, sector, and organization size, analyzing a range of factors related to family status and parenting, looking for a link to women’s lesser representation in top management. But we found no connections. We considered not only whether graduates had gone part-time or taken a career break to care for children, but also the number of times they had done so. We asked about common career decisions made to accommodate family responsibilities, such as limiting travel, choosing a more flexible job, slowing down the pace of one’s career, making a lateral move, leaving a job, or declining to work toward a promotion. Women were more likely than men to have made such decisions—but again, none of these factors explained the gender gap in senior management. In fact, both men and women in top management teams were typically more likely than those lower down in the hierarchy to have made career decisions to accommodate family responsibilities. We even looked at whether simply being a parent—aside from any career changes or decisions related to parenting—made a difference. It did not. Again and again, our core finding—HBS alumnae have not attained senior management positions at the same rates as men—persisted...


At a certain point the belief that a woman’s primary career obstacle is
herself became conventional wisdom, for both women and men. From “opting out” to “ratcheting back,” the ways we talk about women’s careers often emphasize their willingness to scale down or forgo opportunities, projects, and jobs. The very premise seems to be that women value career less than men do, or that mothers don’t want high-profile, challenging work.

Yet framing the conversation like this doesn’t reflect reality—at least not for HBS women, and not, we’d venture, for many other highly educated, career-oriented women...


Companies need to provide adequate entry points to full-time work for women who have, for instance, recently been on a part-time schedule or taken a career break. Our results make equally clear that companies need to move beyond regarding flextime and other “family-friendly” policies as sufficient for retaining and developing high-potential women. Women are leaning in. Most women who have achieved top management positions have done so while managing family responsibilities—and, like their male counterparts, while working long hours. Women want more meaningful work, more challenging assignments, and more opportunities for career growth. It is now time, as Anne-Marie Slaughter has pointed out, for companies to lean in, in part by considering how they can institutionalize a level playing field for all employees, regardless of gender or caregiver status.


So maybe we can work on making sure this bullshit, sexist idea that women "choose" to not advance/get paid as much dies in a fire, like yesterday. It's sexism, pure and simple.
posted by triggerfinger at 6:08 PM on October 21, 2015 [27 favorites]


And yet this year I couldn't get a single female panelist, despite making every effort

I believe it! And part of it may be for the reasons given. (A whole other kettle of sexist fish there!)

However, sitting on a panel can be a golden opportunity to experience first hand, while on stage in front of an audience, forty-five minutes to an hour of the most distilled sexism. I have been interrupted, talked over, condescended to, and incorrectly corrected. All in a set of circumstances where it would have been a severely career limiting move to call any of it out. Or if a miracle occurs and it actually goes well, one of the male organizers will afterwards approach and with a tone and expression of utter surprise, exclaim "wow, you were really good!" And you will muster your last little bit of self control and not punch him in the face. Who wouldn't want to jump at such a golden opportunity?

Fortunately now I'm in house counsel, so the question doesn't come up anymore. Instead I get to sit in the back and take notes. And believe me, if you're a sexist jackass on a panel? Here's one more client you won't be getting, and I don't care if you actually are the best in the field.

So how to translate this to academia? A strong moderator who is aware of the problem and willing to work upstream against it? Not inviting back those who behave badly? All solutions have the same issue- they require a power base to implement, and that is exactly what is lacking.
posted by susiswimmer at 6:13 PM on October 21, 2015 [11 favorites]


Women want more meaningful work, more challenging assignments, and more opportunities for career growth.
This x a million.
posted by susiswimmer at 6:20 PM on October 21, 2015


There was a really interesting analysis of this issue for a conference I am involved with here. The TLDR version is that the majority of reviewers of proposed papers are male, and they then tend to be heavily overrepresentative of certain subcategories of topics that skew more heavily male, i.e. the more hard science end of the scale. They are less likely to be working in the areas that skew more heavily female, e.g. education-related topics, art history, poetry, etc. So the reviewers are just less likely to find papers in those areas compelling, and they score them lower. Even though they don't know (for sure) that the papers are written by women. And then you end up with a male-skewed line-up. So the solution is actually pretty simple: ensure better gender (and international!) diversity of reviewers. Or if you don't want to do it by gender explicitly, at least make sure that your reviewers are evenly distributed over the range of research interests that your conference supposedly caters to.

Finally, we also had an issue with this conference where we had six or seven speakers at the conference opening in a row, who were all male, and we (rightfully) got called out for it. Behind the scenes, though, there were two interesting factors that played into that. First, we didn't select the speakers directly. Rather we asked each one of our sponsors and main contributing organisations to send someone, and they all sent men. Which shows how institutional and structural sexism (senior people in public-facing roles in organisations are mostly men) play into gender representation at conferences. And secondly, in the original selections by these organisations, there were two women (which is still not equal representation, but it's much better). And both of them had to cancel at the last minute. I don't know whether that was just bad luck, or whether it says something about women having to juggle too many responsibilities (e.g. it's entirely possible that one or both of the cancellations were due to carer issues such as a sick child or other dependent). But it also shows that if an organisation is committed to gender equality, it's not enough to put a couple of women in a line-up and hope things work out okay, but you actually have to have back-up plans that involve women too.
posted by lollusc at 6:21 PM on October 21, 2015 [2 favorites]


Oh, and also, barchan's comment reminded me of something similar (but not as serious in scale) that happened to me recently.

I got awarded a grant on a quite niche topic by a specific organisation. Let's call it the Centre for Basketweaving Research. My grant was to carry out research in the area of Painting Baskets with Red Stripes, which is something that no one has thought to try and do before in my field ever. People are painting baskets with stripes (although there's no one doing that at the Centre for Basketweaving specifically), or they are painting them red, so red stripes seemed like a natural progression, but no one else is doing it yet. The grant was specifically for innovative new ideas that brought old ideas together into something experimental and exciting.

The structure of the Centre for Basketweaving is that you apply for grants as the CI on your own project, but with a sponsor who has to be a high-up member of the centre, and the money is administered through them, and you report to them regularly.

Then, six months after I began the work with this new grant money, the Centre for Basketweaving Research decided to hold a symposium on Painting Baskets with Stripes. They wanted to showcase new and exciting experiments in the area of Striped Basket Painting. The symposium was being held within driving distance of my house. Not only did I not get invited to speak, but they didn't even tell me about it in sufficient time that I could attend. (There was an after-thought email to all subscribers to the Centre's mailing list a few days before the symposium, but I guess otherwise they disseminated info via personal invitations and word of mouth).

I was a little annoyed, mainly because I couldn't arrange at such short notice to get there and hear about all the other potentially relevant basket painting research that is going on. But then, I discovered afterwards that they had actually devoted a session to talking about my research that I am doing with the grant they gave me. But they had asked the (male) sponsor of the grant to talk about it instead of me. I have no idea what he would have said, since I had only sent him a preliminary report on what I'd done in the first month or two of the grant, and I hadn't yet updated him on the progress I'd made since then.
posted by lollusc at 6:35 PM on October 21, 2015 [20 favorites]


As a woman in a fairly male-dominated area of STEM, stuff like this happens all the time:
- men talk over me
- men interrupt me while I'm speaking
- men make the same point that I made two minutes ago, but whereas I went unnoticed, they are applauded for their brilliant insight
- when I ask a question in a small-group setting (2~4 people), the person answering looks at a male colleague while replying

I used to blame myself for this: is it because I'm a junior person? Is it because I have a passive personality? Is it because I have a small, non-dominating stature? I worked on this a lot over the years, and these days I'm known as being reasonably direct and assertive, because I thought it was my fault. But now that I'm older, I see stuff like this happening all the time to senior, well-established, firmly spoken women faculty. Nowadays, I see it for what it is: institutional sexism.

One of my advisors had a story about her former student (male). She works in area ABC; she is married to another member of the faculty, who also works in neuroscience, but in a completely different area, DEF. Her student had a question about topic ABC. Rather than ask her, though, he chose to ask her husband! Who doesn't work in ABC!

Also, I just remembered an Anne Churchland story. Soon after she was hired as a new faculty member at Cold Spring Harbor, she ran into an older, male faculty member in the lab building. She recognized him; he didn't, what with her being new and all. He asked her, "Oh, are you the new babysitter?"
posted by tickingclock at 7:10 PM on October 21, 2015 [9 favorites]


Wow, I haven't done anything probably as high-profile as lollusc and barchan, but I've done some very in-depth research in very specific and kind of obscure areas in my field; to the point where I consider myself not only an actual expert on these few things, but I also suspect I'm one of only a few people in my field to have this level of expertise. I suspect this because once I hit a certain level of knowledge I could no longer find other people who were talking or writing about it in the kind of detail I needed. I also talk to other smart people in my field about these things - other actual experts - and they don't have the level of knowledge I do.

I say this because I've been working on various little side projects to distill all the info I've gotten into high-level and succinct summaries for my colleagues and friends in the field. Not because anyone asked me to, but just because I love sharing information and making things more accessible for people. But after these stories, I'm thinking twice about that because wow, would I be pissed if I were to find out one of my obscure topics became the subject of a seminar, journal article or conference talk when I put so much time into pulling so much info together from various sources because there just wasn't very much out there at all.

I mean, I already get that kind of thing on the regular anyway, where I find myself Being Told About Something and I'm like - Wow, I'm certain I was telling you this exact thing days/weeks/months ago. Did you really not hear a single word I said?
posted by triggerfinger at 7:13 PM on October 21, 2015 [6 favorites]


Oh, there was this talk at NYU a few weeks ago:
"Cultures of Genius: Why women are underrepresented in certain academic disciplines"

Sarah-Jane Leslie (Princeton)

Some academic disciplines have significant gender gaps (e.g., philosophy), while others do not (e.g., molecular biology). Sometimes, the phenomenon is characterized in terms of the natural sciences/mathematics having large gender gaps, and the social sciences/humanities having small or no gender gaps. This is a crude characterization, as reflection on disciplines such as philosophy vs. molecular biology illustrates. Are there any general, isolable factors that predict the occurrence of gender gaps across all academic disciplines, and also within the broad domains of natural sciences/mathematics, and social sciences/humanities? My collaborators and I have found that one such factor may be academics' beliefs about what is required for success — in particular, the extent to which innate, immutable, natural talent is emphasized, at the expense of hard work and dedication, predicts the presence and extent of a discipline’s gender gap. In this talk I present the original data in support of the hypothesis, along with several other in-progress studies, which shed light on how these beliefs are communicated, the mechanisms by which they operate to discourage women's participation, and their developmental trajectory. Our findings concerning the latter point suggest that even very young girls are vulnerable to being discouraged by such messages.
I'm not in NYC anymore, so I couldn't go, but it sounded very interesting. I'm leaving it here in case MeFites want to look up her work.
posted by tickingclock at 7:17 PM on October 21, 2015 [4 favorites]


One of my favorites is the Rebecca Solnit essay, now a classic Men Explain Things to Me.
posted by 3491again at 7:38 PM on October 21, 2015 [6 favorites]


It's funny (in the not-funny kind of way) how many people will cycle through any number of somewhat or very implausible theories (such as women not working very hard) in order to avoid the basic explanation of sexism.
posted by Dip Flash at 8:11 PM on October 21, 2015 [1 favorite]


For record I've found the mentioned article - p. 48–49 of an interview.
posted by polymodus at 8:19 PM on October 21, 2015 [1 favorite]


Little surprised by the pushback on the applicability of the math.

First, the assumptions are very simple, but one of my rules of thumb is that simple models are preferable to complex models for an estimate with many unknown mechanisms--simplicity is a feature, not a bug. You can make a million complex models and choose (consciously or not) the one that tells the story you want; if the complex model doesn't match the effect seen in the simple model I'm always skeptical.* Second, more complex models that he could have used ("speakers are not independent" or something) still point to a non-random effect so the idea of a sophisticated model that "corrects" for an old-boys network, shared implicit biases among selection committee members, or whatever, but the overall conclusion would still stand.

Finally, this is a case where the known mechanisms of bias aren't mysterious--there's just a million measurements repeatable experiments where we can isolate and demonstrate them (such as the CV experiments mentioned in the article) or see them in other professions (such as the drastic increase in female musicians in symphonies once they go to blind auditions).

So, it's like we have these individual observations that you'd expect to have an effect, and then we see that there is in, fact, an aggregate effect, I'm not sure why you want to start by assuming some subtle, hidden effect that we haven't identified.

It'd be like hearing forest fires correlate with lightning strikes, and wondering if the data is valid because maybe it means hikers are more likely to light a campfire in bad weather.

*Obviously this doesn't always work, but I'll not even on extraordinarily complex things it's often not bad. For example, with climate change the simple estimate--based on only CO2 infrared absorption and done by Arrhenius over a century ago--you get the right sign and the right magnitude.
posted by mark k at 9:55 PM on October 21, 2015 [9 favorites]


You know, if you bring up overfitting, I'm allowed to bring up underfitting. I'm saying this model underfits, by a fuckton, in the same way that you can't just use a bunch of splines for machine translation. The poking you can't do to the independent model here is qualitative and not too bad in this specific instance, but in other instances you can get vast failures of central limit theorems and lots of statements which are basically sophistry.

I tend to believe that profound, long-range, devilish confounding factors (like the anecdote about the confounding factors of the rat whose maze corridors have to be put in sand) are as common as muck in basically everything that's proven difficult in our knowledge about the world, especially about one another. I don't think it's a paranoiac point to take.
posted by curuinor at 11:03 PM on October 21, 2015


What? I do not understand what you are saying, curuinor. At all.
posted by misfish at 11:34 PM on October 21, 2015


Overfitting is the phenomenon of choosing between models in the situation of complexity mark k mentions. Underfitting is the opposite phenomenon. Have a picture.

The appropriateness of the strength of models in the complexity that they have depends upon the number of parameters to be fit. The IID model claims that one parameter is the one to be fit here: the fraction of women in math. Of course, nobody would claim that this is the complete model of reality. I am arguing that the fact that it's such an impoverished model of reality is very, very important.
posted by curuinor at 11:40 PM on October 21, 2015 [1 favorite]


Why?
posted by misfish at 11:49 PM on October 21, 2015 [1 favorite]


It gets the right answer for the wrong reasons. Would the model be satisfied if the math panel organizers decided to randomly pick women without reference to skill in the topic in the most tokenistic way for the panels?

OK! You see that the IID model is merely very specific, you can leave off of it for predicting the actual behavior of actual people, tokenism is not acceptable and we can use our common sense to figure that out. But that does not mean you leave off of models: it means that you switch to an old one, groaning in recondite complexity, born from before the day of your birth, informal, nigh-unreplicable, the grand model you have of How Other People Act. It's that model we should be working off of! So don't pretend you're not working off of it, if you are. If you formalize, formalize that, instead of flipping coins.
posted by curuinor at 12:09 AM on October 22, 2015


It gets the right answer for the wrong reasons.

I am going to assume that your above arguments about the appropriateness of the model used are in good faith.

However, I will call you out on doing a Thing that is going to earn you a lot of side eye. And that is that whenever a discussion of potential sexism and its impacts on women is occurring, there are always folks who want to say that the methods weren't rigorous enough, or the tone used wasn't respectful enough, or some other distraction from the main point.

So taking your argument that there isn't sufficient complexity and running with it, we can add in various nodes that might be related by three or four different pathways. That women make conscious choices to leave. That they will not make sufficient sacrifices to advance their careers. That they are not as innately talented as the men. That they are not as motivated as the men. Etc etc ad nauseum ad infinitum.

But the tricky part is, these are all symptoms of an underlying cause. You can read up on studies on any of the following: bias against young girls in primary school participating in science or math classes; number of female vs. male athletes in middle school vs. in college; the second shift; the symphony example cited above; the HBS study on the evaluation of the same resume with a 'female' name vs. a 'male' name. The list goes on and on. Or read any of the anecdotes above.

So all of those other 'nodes' do have one common interrelating factor: sexism. They are symptoms as well, not causes in and of themselves. That's the noise in the floor that the 'rats' hear that keeps causing them to make left turns off the career highway.

And after all the common terms are dropped, and the things multiplied by zero are removed, one is left with the math proposed in the FPP. Only the numbers used are wrong. Because it should be 50/50.
posted by susiswimmer at 12:46 AM on October 22, 2015 [15 favorites]


I am arguing that the fact that it's such an impoverished model of reality is very, very important.

You're really not. You're making theoretical statements divorced from the subject matter at hand. The model says it's not just random chance, the interview says the non random part is bias (and generally implicit bias.) And you apparently agree with both statements, so what's "important" let alone "very, very important" that we're missing? Discussing alternate modeling approaches that might distinguish between tenure committee bias and citation rate bias is apparently a preference you have, but is a bit beside the point.

You might want to reread the original interview, which is actually about women in mathematics and why they are underrepresented, rather than about strengths and weaknesses of statistical methodologies. Indeed,
Changing from the real-life situation to a probabilistic model is quite an imperfect process, and people can rightly criticize details of that change to the point where the underlying message is obscured.
posted by mark k at 12:52 AM on October 22, 2015 [15 favorites]


One thing I really dislike when talking about sexism on the internet is how if we aren't careful with our words we tend to unconsciously prioritize "being the most correct" over "not coming off like an asshole".

I really really dislike that A LOT.
posted by Annika Cicada at 7:23 AM on October 22, 2015 [4 favorites]


Women want more meaningful work, more challenging assignments, and more opportunities for career growth.

Yes. They are not all caregivers who want flexibility, especially after doing something like an HBS MBA and it's crazy to keep thinking that way.

I am in an incredibly gender balanced department with a woman leading it. I am beyond fortunate not to be mommy-tracked, mostly due to having a mom in charge and lots of working moms who still are devoted to work 24/7, and my repeated, loud assertions to anyone that will listen that my husband is the caregiver and I don't need flexibility (I wouldn't mind it, but am incredibly conscious of the perception of not leaning in, so I lean in with overkill). On top of all that I am incredibly good at my job, in all quantitative and qualitative factors that matter. And, in the opinions of everyone I have talked to, head and shoulders above my peers at my level.

So, despite all of that - I was recently passed over for a promotion (confounding in many levels, particularly because it had been mentioned to me as a possibility for literally years) and told that I needed some additional experience in X, and now I had the "opportunity to really step up!" I got the feeling I was supposed to be grateful for this opportunity to demonstrate (once again) how incredibly good at my job I am, of course, with no raise or title increase. While I don't think overt sexism was at work, it's hard to think it's not a factor at all when there is someone two levels above me with far less experience in X than I have, same age as me, same educational background, but who happens to be a good-looking, tall, male.

To be honest, this is a fairly small story in my career and I have experienced dozens that are similar (how about doing the job I was interviewing for while my boss continued to interview less qualified men and then told me the "decision was very difficult"!?) there are just countless microagressions on a path to "top of your field" and it's just so ridiculously exhausting that at some point most of the women I know just ask themselves what the point is? If I wasn't the breadwinner and stubborn like a pit pull I would probably quit and do my own thing just to avoid the constant bullshit and perception you are "doing it wrong" no matter what you do. I just use the anger to fuel my drive to prove them wrong, but not sure that is healthy either.
posted by rainydayfilms at 8:00 AM on October 24, 2015 [3 favorites]


The American Mathematical Society publishes statistics on women mathematicians every year. The most recent stats show that a special session with no female organizers will be on average 81% male, while a special session with at least one female organizer will be 73% male.
posted by yarntheory at 10:08 AM on October 24, 2015 [1 favorite]


« Older I Like Big Books And I Cannot Lie   |   The Color of Debt: How Collection Suits Squeeze... Newer »


This thread has been archived and is closed to new comments