ABC News reports on a new study by researchers at the University of Chicago and Northwestern University which supposedly determined that women with higher testosterone levels "take more risks and are more likely to choose a finance career." Testosterone levels made no difference in men's career choices in the study. The sample size was 500 graduate students.
I know that hormones play a real role in our behavior--no matter our sex--but these kinds of studies worry me. It feels as if isolating only one factor like this, especially one so biologically-determined, underplays all of the other huge influences on how we choose careers, get educated, seek mentors, develop an identity, determine our own gifts etc. The socialization, for example, that we experience as a result of our socioeconomic class, seems like a far greater influence on whether we see ourselves as "fit" for a career in finance, than whether we have a slightly higher testosterone level.
Which is all to say--okay, do the hormonal studies, but don't forget to couch them in the context of what I see as far more powerful social, economic, and psychological factors. Your thoughts?
0 TrackBacks
Listed below are links to blogs that reference this entry: Does Testosterone Play a Role in Career Choice?.
TrackBack URL for this entry: http://www.feministing.com/cgi-bin/movabletype/mt-tb.fcgi/15701
I'm not buying it. Risk taking is relative. Everyone takes what he or she perceives to be a risk. I wonder if it's more of a placebo effect, if anything, since testosterone is tied in to being assertive and dominant.
My thoughts exactly. The act of "choosing a career in finance" can be seen my some as a pretty non-risky choice-- i.e., you've gone ahead and picked something that generally has a whole lot more of financial security and social acceptance than, say, being a freelance writer or community organizer. I work in finance (as a lawyer, yes, but still working on large corproate transactions where there is time pressure and money at risk if a comma is misplaced or a document does not get done in time) and I know I got here by following the path of least resistance!
I think the real problem here is the way we characterize the risk/difficulty profile of different jobs based on who tends to do them and how we stereotype those people and value their time. If you really think about it, is a 20 year old single dude who drinks hard, works 20 hour days at the office and trades securities on Wall St really working harder, under more pressure, and taking more risks than (totally random example) a 35 year old single mom who works, raises her kids and studies to prepare for a second career? Having been a mom (even under ideal wealth/partner circumstances) and a finance worker, I will tell you the answer to that question is most assuredly no.
I agree.I dont think risk is based on gender. I also definitely think socialization has something to do with it.
It was interesting that some women had as much or more testosterone in there saliva samples. I didnt know that was possible?
Except, for a placebo effect to occur, the person actually has to *know* they have consumed/have the substance in question.
People don't actually know their testosterone levels. Its not a part of any regular medical check-up, and only maybe 1 in 1000 people knows what their testosterone level is.
How could a person be psychologically affected by having a higher-test (higher expectations for self based on test), if they don't even know they have it?
Ponder that for a second.
It seems as though it's in the study itself - while testosterone is a statistically significant covariate, it only accounted for 5-6% of the difference. Other factors accounted for 94% of the difference.
I think this 101 is relevant: http://finallyfeminism101.wordpress.com/2007/05/10/faq-but-men-and-women-are-born-different-isnt-that-obvious/
In other words biology plays some small part in behaviour but it is dwarfed by the enhancement of those effects by socialisation. Studies that attempt to quantify what the real biological difference is can only aid us in pointing out the extent of socialisation.
Hi there. I find maths and statistics difficult, so I could only get the gist out of the article. Can I ask how you drew that conclusion from the article, because I couldn't see it (also reading things as iddy-biddy PDFs might mean I might actually not have seen it)?
As a neuroscientist, I can appreciate the attempt to isolate single factors for study. I think most of us realize the greater complexity and in a lifetime of study hope to understand it better, but complex study designs are most often poor study designs. As kat mentioned in a prior post, testosterone only accounted for about 5 percent of the variance. I think the media hype over the study's significance is the bigger problem. Scientific journalism is generally pretty poor these days.
I agree with monkey_doc here - a lot of the issue is with how it's reported (see the use of words when describing the behaviours of sperm/ova for some really fascinating gender stereotypes coming out on a cellular level).
There are theories out at the moment that testosterone works more like adrenaline than estrogen - it may well be mostly responsive to stimuli, rather than consistently in one's system, which changes one's behaviour. I'm sure that the scientists out there can find some sources on this?
"that testosterone works more like adrenaline than estrogen "
Great! Trading one sexist theory for another. I can see how the scientists and society will use that one against women!
Perhaps mindprovender was referring to the idea of activating vs. organizing effects of hormones? For the lay audience, the first refers to temporary hormonal effects due to fluctuating levels of the hormone. The second refers to permanent effects of exposure to the hormone, usually during developmentally sensitive periods. I don't see the sexism in that. Estrogen and testosterone each have both effects. If that's the sense mindprovender was referring to, then the speculation is wrong in the sense that neither estrogen and testosterone is quite like adrenaline, which I think has only demonstrable activating effects (I could be wrong).
Sorry, my brain fell out of my skull.....I'm lost?
I'm a media and bus major, not science.
The study has already been deemed laughable by many other scientific groups. It's the chicken vs. the egg problem, women in (medically defined) high stress situations have higher levels of testosterone than those who are not, the stress generates it, and the process feeds on itself.
Women who play a lot of sports have much higher levels of testosterone, it's what gives them their incredible physiques. Are they more likely to enrol in an MBA programmes? Anecdotally,I'm not aware of any studies, I would say no, their major distribution mirrors the one for the general female population.
This. The study needs some Anne Fausto-Sterling-style analysis, methinks.
There is no more chicken vs. egg problem. It has been shown that situations affect testosterone level. I use "situations" to encompass stressful situations, sexual situations, competitive situations, etc. All of these can either increase or decrease testosterone levels. That is different from the amount of testosterone which a given individual is able to produce or absorb. Think of a baseline of testosterone with situationally induced fluctuations. The process doesn't feed on itself b/c the fluctuations are transient.
Even if this study were credible, I see it as a tool for naturalizing the wage discrepancy. I would be interested in how they define "risk".
As I see it, it is LESS risky to go into a more lucrative field, and more to go into traditionally female-dominated ones, which are less values and poorly compensated.
I think this is junk science. For years "scientists" have been claiming that women are less likely to take risks. Now they're saying that testosterone levels contribute to risk taking. All of this denies the ways in which socialization impacts our behavior. To pin it all on hormones is too simplistic.
Are you saying this is junk science because you've read the article and have identified a critical flaw in the study design, analyses or conclusions? Or are you simply dismissing it because you don't like the finding? The latter course is a dangerously anti-science stance (accepting only science you find personally acceptable).
I haven't read the article - I'm simply keeping an open mind. The lead author is a professor at the University of Chicago - it would hardly seem appropriate to use mocking quotations on the word scientist to refer either to him specifically or scientists in general who study steroid hormones. Aside from this article, the scientific literature on testosterone and risk-taking behavior is fairly substantial. No one I know among my scientific colleagues would assert a 100% directly causal relationship. To do so would be ludicrous. But neither do the authors of this article. OTOH, to state that hormones have no influence on risk taking is about as intellectually defensible as creationism.
Similarly, in response to the person (not Serena) who asserted that many other scientific groups have deemed this study laughable - can you provide evidence to back this assertion? The study just came out.
If we're being intellectually honest here, the main problem is not with the science itself, but with the media's characterization of it. Again, I'm not impressed by the hyperbole in the media. I doubt the hype is in the article, but I'll read it just to make sure. Scientists are sometimes guilty of hype as well, especially when they get media attention/fame from their studies.
And to add to what monkey_doc said, how do we know what the scientists 'wanted' to show when they started the study? A good scientist will report what they find no matter whether it matched their expectations or not. They will then attempt to interpret those findings within the framework of literature already recognised.
Yes, GOOD scientists do this. It wasn't long ago (and probably still happens now, though I can't cite anything right off the top of my head) that some scientists just ignored evidence that did not fit into their expectations or their hypotheses. I tend to think, though, that monkey doc's logic is more what happens now - sometimes a study is done well and everything is presented but then the media spins it to increase interest and sell newspapers, magazines, etc.
Yes, well scientists have to pass peer review to get something published, journalists just have to get it past their editor, and anyone can post something on the internet.
I don't think there should be an inherent antagonism between feminism and every form of scientific reductivism, reducing a complex situation into a set a core principles or relationships is the bedrock of science, but there still is a lot of gender typing in this study that needs to be called out.
Like "Risky Behavior", what the hell is that? Are sex-seeking women considered 'riskier' than sex-seeking men? The study used a gambling game to calculate that, but generalizing from gambling to other 'risky' behaviors involves a substantial leap of logic. Also, it's complete bullshit to conclude women seek the financial sector because it's risky, it's much more accurate to say the financial sector rewards high testosterone, which a study of top male investors has already argued.
Honestly, I think the controls on this study are insufficient to argue for the accuracy of the conclusions without a much larger and diverse sample (i.e. more than 140 women). I think the publicity of this study may have taken its conclusions beyond the scope that the authors intended, but whatever form it takes, pop science is a breeding ground for patriarchal bias.
And I would argue that the gender stereotyping is not actually IN the study. It is in the media response to the study. I've now read the article and it seems pretty neutral to me. The authors end with a good cautious recommendation for future research: "Future studies should also address the interplay of biological and sociocultural factors in the emergence and maintenance
of between- and within-gender differences in financial
decision-making and other types of risk behavior."
Note also that the authors do not generalize across risk behaviors - they chose gambling because it is logically related to financial success, both in Vegas and in the stock market. They never measure, speculate or otherwise comment on sexual risk-taking behavior, in women or men. That's all been in the reporting on the article. When they mean risk-taking and risk-aversion, they are specifically referring to financial risks.
"If we're being intellectually honest here, the main problem is not with the science itself, but with the media's characterization of it."
Really? And that would be because science is so clearly uncontaminated by ideological biases that shape its guiding epistemologies in fundamental ways it cannot see?
While I cannot, of course, speak for the poster Serena, I interpret her objection to the study not as what you caricature as "a dangerously anti-science stance," but rather as an expression of the entirely valid concern that, wearing the cape of "objectivity," the scientific community has always, and to some extent continues to, reify what is merely hegemonic ideology (gender, racial, etc.) as biological reality.
While the following are the words of someone replying to your comment, they are keeping with the spirit of your own. A sentence like - "A good scientist will report what they find no matter whether it matched their expectations or not" - belies a deeply worrying absence of familiarity with the extremely broad, well-articulated, and internationally-recognized body of scholarship on gender and science of the past three decades, as well as a more general, though no less alarming, ignorance of the existence of ideology and/or how it operates in the human mind.
If we were actually being "intellectually honest" here, we would be including in our dialogue a meaningful engagement with the critique of scientific practice, not demonizing and dismissing it out of hand.
I hear what you're saying. I don't object to an honest dialogue about biases present in science. They are clearly there and I acknowledge the truth of your assertions about how those biases affect the literature. If that means to anyone that any scientific study that finds a sex and/or gender difference in biology or behavior must be, ipso facto, incorrect, without even reading the study, then I think its fair to label that a dangerously anti-scientific stance. I do object to decrying an article to be junk simply because the data and its conclusions may not fit comfortably into a particular world view - that was my interpretation of Serena's comment. It may not be the correct interpretation, but I think its a reasonable one, particularly given the choice to use mocking quotation marks on the word scientists. Finally, I do not say that the article is without flaw - simply that what people are seemingly objecting to here are characterizations of the science that appear only in the media, not in the article itself.
Scientists are far more aware than the general public of the infulences of ideology on opinion. The history of science is littered with examples of how scientific break-through was only possible because someone looked objectively at the results of an experiment rather than being swayed by established doctrine. Science students are familiar with many of these stories.
Now, it is true that if you obtain a result that doesn't seem to fit with established knowledge your first thought is to doubt yourself. But you don't throw away your result, instead there is a process of careful investigation to confirm its meaning and authenticity. It is this process that the peer review procedure predominantly checks for.
Being a scientist, I can tell you that I have been through this process first hand and that science isn't as easily corruptable as you seem to think. Critical thinking is at the forefront of the process, and ideology should play no part. A good scientist is at pains to put aside their preconceptions - far more than any average member of the public - this is what they are trained to do.
I don't disagree that there has been trouble with the scientific treatment of gender. However I would argue that it has only been trough scientific debate that misconceptions have been able to be addressed. Furthermore, the rigour of the scientific process that demands that assumptions and procedures be articulated has been instrumental in improving our understanding of the realities to this point.
To dismiss a piece of science on the basis that science has been wrong in this area in the past is irresponsible. Each published result needs to have its merits examined within the context of the validity of the assumptions that underpins it, the procedures followed and the knowledge established by previous literature. Science is self-correcting. A bad piece of science cannot stand for long.
You have assumed that I'm not a scientist (is that because I was being 'critical' of science?); I am, so am well aware of both the history and first hand praxis thereof. "Corruptability" isn't the same thing as interpellation, and while the importance of extricating preconceptions from process is indeed part of our training, that is a concept in no way related to the phenomenon of ideological bias that I was alluding to in my post.
"Science is self-correcting. A bad piece of science cannot stand for long." I suppose then that you and I have a different definition of the word "long"; I consider several centuries to be rather a long time indeed. (See, e.g. The Gender and Science Reader, ed. Lederman & Bartsch). Such a sentiment is also governed by the inherently flawed - and widely damaging - belief that science can attain complete objectivity. It's been said many times, many ways, from Derrida, through Bordo, through Fausto-Sterling, through innumerable others, but apparently I have to repeat it yet again: Peer review is in no way capable of ferreting out actual ideological bias (which is not the same thing as "opinion" or "preconception", both of which are conscious processes).
I have tried for the greater part of two decades to bring the to some extent dichotomous entities of "the sciences" and "the humanities" into a meaningful collaboration; in whatever medium, from the colloquial (here, and other blogs) to the professional (academe), the results are always depressingly the same, with each side clinging vehemently to their biases and interested only in defending the rightness of their position, not in re-examining it in the interests of the greater good (which is what each side claims to be interested in).
"You have assumed that I'm not a scientist (is that because I was being 'critical' of science?)"
Well, it was more the way in which you were being critical of science that drew my assumption (I think monkey_doc elaborates what I mean best). In general I tend to assume people are not scientists because we are a minority of the population - this also affected my assumption.
"I am, so am well aware of both the history and first hand praxis thereof. "Corruptability" isn't the same thing as interpellation, and while the importance of extricating preconceptions from process is indeed part of our training, that is a concept in no way related to the phenomenon of ideological bias that I was alluding to in my post."
Here I disagree with you. Ideological bias is a form of preconception.
" "Science is self-correcting. A bad piece of science cannot stand for long." I suppose then that you and I have a different definition of the word "long"; I consider several centuries to be rather a long time indeed. (See, e.g. The Gender and Science Reader, ed. Lederman & Bartsch)."
I suppose what I should have said is that, bad science cannot stand the tests of mature science long when it is thoroughly investigated. Social movements can be an impeutus for thorough ivestigation of an idea.
"Such a sentiment is also governed by the inherently flawed - and widely damaging - belief that science can attain complete objectivity. It's been said many times, many ways, from Derrida, through Bordo, through Fausto-Sterling, through innumerable others, but apparently I have to repeat it yet again: Peer review is in no way capable of ferreting out actual ideological bias (which is not the same thing as "opinion" or "preconception", both of which are conscious processes)."
I don't know if I believe that science can attain complete objectivity but I would argue that the process of science brings it closer to that ideal than any other methodology.
What Peer review can do is point out whether the assumptions that underpin a work are appropriate or justified. So if we take this study for instance, a reviewer could have said that testosterone in saliva was not a good measure of testosterone's long term effects because it can be raised by stress. The researcher would then have had to seek out a measure that was indicative of long term testosterone levels (finger ratios) or withdrawn the conclusions to do with long term effects from the paper.
"I have tried for the greater part of two decades to bring the to some extent dichotomous entities of "the sciences" and "the humanities" into a meaningful collaboration; in whatever medium, from the colloquial (here, and other blogs) to the professional (academe), the results are always depressingly the same, with each side clinging vehemently to their biases and interested only in defending the rightness of their position, not in re-examining it in the interests of the greater good (which is what each side claims to be interested in)."
I also happen to be a published writer of fiction and poetry. I've organised cross-over events like science in art shows, science fiction writing competitions etc. (Did you assume I had no interest in the humanities because of my defence of science?) I'm aware of the importance of the different disciplines of study. I'd posit though, just on the vibe I get from your comments, that I take more of a 'salad-bowl' approach compared to your 'melting pot.'
There are mechanisms within science to challenge established lore, I see the role of the humanities (in this context) as providing the impeutus to initiate those mechanisms, to begin studies to investigate the assumptions and (if correct) to overturn that established lore.
"Here I disagree with you. Ideological bias is a form of preconception."
That's not something that can be disagreed with, anymore than you could "disagree" with me that when I use the phrase 'darling bulldog' I'm referring to my best friend. (Which is to say, you certainly could 'disagree' with that fact, but it would be meaningless and absurdist from a communicative standpoint). An essential part of the problem here is that you're not familiar with the concepts in play. When I, or Bordo, or Derrida et. al. use the phrase "ideological bias" we are referring to a very specific thing that generally bears little relation to lay understandings of the term. The vast majority of scientists whom I've met, worked with, spoken to, etc. have at best a rudimentary conception of it, and, given the way science education is still structured, that is unlikely to change. Which in turn means that reification will continue, which in turn means that human lives will continue to be negatively impacted by a science that is acting to some degree as an (if you like, "unwitting") instrument of sociological hegemonies.
"What Peer review can do is point out whether the assumptions that underpin a work are appropriate or justified."
Sigh. You follow this assertion up with physical examples (how testosterone fluctuates, etc), when the whole point is that peer review cannot effectively discern whether the __onto-epistemological__ assumptions that underpin a work are appropriate. I'm not suggesting PR never has, but it is, by definition inherently prone to the reverse: remaining blind to the ways in which ideological bias (i.e. the, for simplicity's sake let's call it 'invisible' variety, as vs. the visible obvious type - e.g. a known racist who sets up experiments to prove the inferiority of black people) structures every single dimension of research (including that which is most often mistakenly thought to be the most 'innocent' - the moment of idea conception). Because this is so, not every bias can be corrected by mature science, if only we give it enough time.
"I also happen to be a published writer of fiction and poetry. I've organised cross-over events like science in art shows, science fiction writing competitions etc. (Did you assume I had no interest in the humanities because of my defence of science?) I'm aware of the importance of the different disciplines of study. I'd posit though, just on the vibe I get from your comments, that I take more of a 'salad-bowl' approach compared to your 'melting pot.' There are mechanisms within science to challenge established lore, I see the role of the humanities (in this context) as providing the impeutus to initiate those mechanisms, to begin studies to investigate the assumptions and (if correct) to overturn that established lore."
Ah, but I wasn't bringing up the issue of collaboration with the implication that you had no interest in the humanities - nor simply to champion "the importance" of the different disciplines. What's at stake is the very nature of interdisciplinarity; what it is, how it works, why it's vital, etc. Unless and until the scientific community is willing to truly interrogate its prevailing epistemological operations -- rather than, as your scenario suggests, a relationship in which the humanities sends a telegram over to the sciences ~ 'Hey Science, lore X looks suspicious, you should check it out' ~ and then they rely on their own internal existing mechanisms to investigate -- the status quo is essentially assured.
While there are few things I find quite so ethico-politically objectionable as the melting pot concept, if, in this particular context, your 'salad bowl' approach retains the separatism of thought itself, it could never hope to free science from the bias that ultimately takes lives and destroys eco-systems, and therefore would be something I could never endorse. It seems we must agree to disagree...
I'm interested to know more by what you mean by "ideological bias" then if the definition is different from the dictionary one.
To me an ideology is an idea that one might base their behaviour or world view on. They may or may not be aware that they hold the ideology. Yet it could be uncovered by logical argument or by examination of the way an experiment is set-up, etc. so long as someone is aware and thoughtful enough to examine the possibility. This may not occur immediately but could be done at a later date, so long as the details of how the experiment were conducted were sufficiently detailed. If the person who set up the experiment was biased, then when this bias is assumed the results should be able to be reconstructed in that light.
It seems we do have to disagree though. I don't think the scientific process is broken. I do think that some of the formalism surrounding how we practise science in the present age needs reworking though.
It aye helps to actually look at the study in order to form conclusions...
And now that I have read it, I can say that although I agree with some of the commentary in the link you provided, some of it was overly dismissive. For example, the concern over the risk measure - measures like that are quite commonly used and generalize pretty well. I cannot speak to that particular measure, but gambling tasks are very good measures of risk aversion and they are actually real world tests, or at least as close as one can come in a lab setting.
The authors should have reported data from their other measure of risk aversion (the Arrow-Pratt), instead of simply asserting that the measures produced the same statistical result. The digit measurement (ratio of the index to ring finger's length) is well accepted as a measure of prenatal androgen exposure. I think the Reading the Mind in the Eyes test (a social cognition measure I use in my own autism research) is not as good for its relationship to prenatal testosterone, but its always good to try to have a backup measure for each construct in your study. Note too that the Eyes task was not related to career choice, while the digit measure was.
I agree, however, with the media's completely over the top and unhelpful hype about the study. The blog you provided nailed that completely.
It's possible that what Alex Catgirl asserts is true:
" It's the chicken vs. the egg problem, women in (medically defined) high stress situations have higher levels of testosterone than those who are not, the stress generates it, and the process feeds on itself."
For example, another recent article asserts, among other things, that marriage seems to decrease testosterone levels in men:
http://www.newscientist.com/article/dn17692-too-much-testosterone-disrupts-family-life.html
So, it's possible that women also can also change their testosterone levels in response to stimuli.
The study in question does prove that the testosterone level itself partially affects risk-taking behaviour in women. It's still a significant finding; it shows that our hormones correlate to our behaviour. Whether it is, as you say, chicken or the egg, is not yet proven.
A study that tracked women before they entered the financial industry, and followed up with them during their careers, would settle the question by providing a baseline testosterone level to compare to that of a woman during her career. But until that's done, both the nature and nurture hypotheses should be considered valid.
I've been immersed in the feminist blogosphere for most of this year. It's challenged and changed my assumptions quite a bit, especially when it comes to the "nurture" side of the gender differences debate. I've come to accept that socialization plays a large role in the development of gender differences. Feminism has even given me the courage to pursue a second career in nursing.
But as a scientist first, what I can't get on board with, however, is the visceral negative reaction to any scientific evidence that disproves the idea that gender is purely a social construct. And then every once is a while, someone on here wonders why more people (of both genders, but especially men) don't call themselves feminists...
Just to be entirely clear - I don't disagree with the idea that socialization is important. No one should read into my defense of the study that I believe that biology is destiny - I don't. I'm not interested in the nature vs. nurture false dichotomy either. I'm only interested in the misattributions in the commentary on the article. I think we can distinguish between what the article actually asserts and what the media and lay audience responses have been. Nowhere in the published article do the authors suggest that testosterone is the only important influence in risk taking. They in fact suggest that future studies should look into androgen interactions with sociocultural factors.
The issue in this study was whether testosterone levels in men and women could predict risk aversion (yes, especially at low levels for both sexes) and career choices (yes, but only for women with high levels in this sample), not whether testosterone levels fluctuated in response to certain stimuli (other studies show they clearly do). The authors recognized that a single measure of salivary testosterone would prove vulnerable to that problem, so they chose a proxy measure of prenatal testosterone exposure, the index to ring finger length ratio. The fact that this measure also had predictive value is suggestive, if independently replicated, that there are androgen influences that are independent of cultural influences that are important in risk-aversion. It does not mean that sociocultural influences on androgens after birth are not important.
That's a discussion about the study. Here is the Journal Article: http://www.pnas.org/content/early/2009/08/20/0907352106.abstract
You'll need a subscription to the journal (or to purchase the article), otherwise you'll only be able to read the abstract.
I wouldn't be surprised if testosterone levels played a role in career choice. As it turns out, names seem to play a small, but statistically measurable, role in career choice. If your name has an influence, then almost everything would.
I think your fears are justified, particularly because the media and ordinary people like to oversimplify complex arguments because doing so makes it easy to explain and to disseminate.
Socialization plays a huge role but I think one can't discount or denigrate biology, either. The way that things work in combination with each other is still so poorly understood, and our theories change with our understanding of the topic. If we still don't understand 5% of how the brain really works, this underscores just how little we know about anything.
I for one am psyched to read of this study cause the next time my girlfriend chastises me for not wearing my bike helmet I can inform her that as a female-to-male transsexual I am merely a victim of testosterone-induced high risk behavior.
She's only chastising you in the first place because she's a victim of estrogen-induced nagging behavior....
OR having been made aware of the possible nuanced effects testosterone could have on you, you might examine your own decision making in that light, and decide that 'Hey, maybe I should wear a bike helmet.'
The research is not saying that those with high testosterone levels are fated to take risks, just that it is one small indicator/factor.