Friday, November 30, 2018

Do women in physics get fewer citations than men?

Yesterday, I gave a seminar about the results of a little side-project that I did with two collaborators, Tobias Mistele and Tom Price. We analyzed publication data in some sub-disciplines of physics and looked for differences in citations to papers with male and female authors. This was to follow-up on the previously noted discrepancy between the arXiv data and the Inspire data that we found when checking on the claim by Alessandro Strumia and his collaborator, Ricardo Torre.

You find our results on the slides below or you can look at the pdf here. Please be warned that the figures are not publication quality. As you will see, the labels are sometimes in awkward places or weirdly formatted. However, I think the results are fairly robust and at this point they are unlikely to change much.



The brief summary is that, after some back and forth, we managed to identify the origin of the difference between the two data sets. In the end we get a gender-difference that is not as large as Strumia et al found it to be in the Inspire data and not as small as we originally found it in the arXiv data. The male/female ratio for the citations normalized to authors is about 1.5 for both the arXiv and the Inspire data.

We then tried to find out where the difference comes from. This is not all that obvious because the particular measure that Strumia used combines various kinds of data. Eg, it depends on how frequently authors collaborate, how many papers they publish, and how much those papers are cited.

We know that the total number of citations is comparable for men and women. It turns out that part of the reason why women have a lower score when when one counts the total citations divided by the number of authors is that women write (relatively) fewer single authored papers than men.

This, however, does not explain the entire difference, because if you look at the citations per single-authored paper (ie, without summing over all papers), then women also get fewer citations.

We then looked at where those citations are (or are not) coming from, and found that both men and women cite single-authored papers with female authors at a lower frequency than you would expect from the share among the citeable papers. It turns out that in the past 20 years the trend in women-to-women citations (single-authored papers only) has gone up, while for men-to-women citations it has remained low.

It is not a huge difference, but since there are so many more men than women in those fields, the lack of citations from male authors to female authors has a big impact on the overall number of citations that women receive.

In all those analyses, we have removed authors who have not published a paper in the past 3 years or who have fewer than 5 papers in total. This is to avoid that the higher percentage of dropouts among women pulls down the female average.

One of the most-frequent questions I get when I speak about our bibliometric stuff (not only this, but also our earlier works) is what are my own scores on the various indices. I usually answer this question with “I don’t know.” We don’t dig around in the data and look for familiar names. Once we have identified all of an author’s papers, we treat authors as numbers, and besides this, you don’t normally browse data tables with millions of entries.

Having said this, I have come to understand that people ask this question to figure out what are my stakes, and if I do not respond, they think I have something to hide. Let me therefore just show you what my curve looks like if you look at the index that Strumia has considered (ie the number of citations divided by the number of authors, summed up over time) because I think there is something to learn from this.



(This is the figure from the Inspire-data.)

Besides hoping to erase the impression that I have a hidden agenda, the reason I am showing you this is to illustrate that you have to be careful when interpreting bibliometric measures. Just because someone scores well on a particular index doesn’t mean they are hugely successful. I am certainly not. I am 42 years old and have a temporary position on a contract that will run out next year. I may be many things, but successful I am not.

The reason I do well on this particular index is simply that I am an anti-social introvert who doesn’t like to work with other people. And, evidently, I am too old to be apologetic about this. Since most of my papers are single-authored, I get to collect my citations pretty much undiluted, in contrast to people who prefer to work in groups.

I have all reason to think that the measure Strumia proposes is a great measure and everyone should use it because maybe I’d finally get tenured. But if this measure became widely used, it would strongly discourage researchers from collaborating, and I do not think that would be good for science.

The take-away message is that bibliometric analysis delivers facts but the interpretation of those facts can be difficult.


This research was supported by the Foundational Questions Institute.

45 comments:

Pascal said...

>The reason I do well on this particular index is simply that I am an anti-social introvert who doesn’t like to work with other people.

While I'm reading this, blogger is displaying an ad for an anti-depression drug. I don't if that is a coincidence, or another prowess of A.I...

Uncle Al said...

"I am an anti-social introvert who doesn’t like to work with other people" In DSM IV, Asperger's syndrome. Absence of smiles, avoiding another's eyes when speaking, periods of profound manic productivity, comfort in isolation. "The Accountant" (2016).

DSM 5 has "autism spectrum." Being hugely inside yourself sells SSRI brain buckets of ice water inflicting escalating misery and side effects. It is war on fulgent people particularly targeting young males.

No guilt! Maximum gravitation at a solid mass’ surface is not a sphere. One can do 2.6 % better. This pleases me.

Matthew Rapaport said...

You are "not successful" only by a narrow measure of "success" even within what is at least orthogonal to your field. You have perhaps become the Slavoj Zizek of physics! Perhaps no one wants to employ you but you have a huge fan base, your books sell, and there will always be speaking opportunities! 😊😉

Steven Mason said...

First I'll say that I think it's worthwhile to ask questions like "Do women in physics get fewer citations than men?" It also looks like you took care to find meaningful answers (as any good scientist should). I'll respond to two parts.

Sabine wrote: It is not a huge difference, but since there are so many more men than women in those fields, the lack of citations from male authors to female authors has a big impact on the overall number of citations that women receive.

I'm guessing that before you began this project, you already knew that there were many more men than women in those fields. I wonder if there are significant differences in the ratio in various nations. If some nations have significantly more women entering those fields, do we know why?

I also wonder if you think it's important to encourage women to enter those fields, and what kinds of things nations could do to encourage women. Of course, when we talk about "encouragement" for women or men, we'd have to talk about early education and cultural influences.

For most of my life I've wondered about male predominance in mathematics, physics, engineering, and even chess. Some people lean toward nature, some to nurture, and some say it's a combination. "Nature" doesn't necessarily mean less ability; it could also mean less interest, leading to less participation.

I remember Marilyn vos Savant, an American woman who, for a while, had the highest recorded IQ in the world. She also wrote a magazine column called "Ask Marilyn," which was sort of like, "The smartest person in the world will answer your questions." One time a reader asked, "Are you good at chess?" Marilyn said, "I have never played chess. In fact, I don't play any games at all, including card games."

In 2005, the magazine "Chess News" mentioned Marilyn in one of their "Question of the Week" articles. The provocative, tongue-in-cheek question was, "Are men smarter than women?" Here's a blurb from that:

Science, like chess, attracts bright people, and there are many more males in both fields. So can we conclude that men are smarter? Not necessarily, says Marilyn vos Savant in PARADE Magazine. "Chess was developed by males for intellectual sparring with other males. Maybe females simply don’t find the game as fascinating."

Marilyn's remark refers to interest, not ability. But that raises questions. What role does gender play in our interests?

Steven Mason said...

I know plenty of male scientists who have no interest in chess. I used to play chess a lot when I was a teen, but once I reached a competitive level, I discovered that further progress was largely a matter of memorizing hundreds and then thousands of move sequences. I had no interest in endless hours of rote memorization for the sake of a game. To this day, whenever I see two people playing speed chess and taking several seconds per move, I feel pity for them, because they are simply - and literally - going through the motions.

Of course, who am I to say they are wasting their time? I realize my pity is misguided. For similar reasons, I feel pity for all the children who spend countless hours of their precious childhood memorizing word spellings, so they can compete in spelling bees. I'm not proud to admit that I consider the rigors of spelling bees to be a mild form of child abuse. Due to my bias, I can't imagine any child who naturally wants to spend thousands of hours on rote memorization of word spellings. To me it's a red flag for some kind of obsessive-compulsive disorder. I could be wrong: maybe there hundreds of kids born each year who, for reasons I can't understand, feel great pleasure at memorizing word spellings and can't get enough of it.

By now you might be thinking I'm rambling, but I'm still talking about the question of "interest." It's a lot easier for me to understand how someone - male or female - can get interested in science, math, physics, or engineering, than to understand how anyone can get obsessively interested in chess or spelling bees.

Steven Mason said...

Okay, let me move on to the second part:

Sabine wrote: The reason I do well on this particular index is simply that I am an anti-social introvert who doesn’t like to work with other people.

Two responses: First, my gut feeling - which is by no means reliable - is that you are inserting a bit of your humor into this statement. Even if what you say is absolutely true, you're expressing it in a way that suggests self-deprecating humor, at least to my American ear. If I'm wrong, if no humor was intended, than you might be the first person I've ever encountered who could say something like that in a strictly serious manner.

I can't explain why I seem to be so interested in figuring out when you're being funny, but it's probably because I've always been that way with everyone. I'm not a jokester or a prankster, but I really appreciate a sense of humor. I recently heard comedian Stephen Colbert say, “Do you know what I like about comedy? You can’t laugh and be afraid at the same time. If you're laughing, I defy you to be afraid.” To that I would add that good humor requires good, clear, critical thinking. In fact, just this morning I told my wife, "If I couldn't laugh at Trump, I'd be depressed." And I meant it.

Second, I'm curious to know, if you really are an anti-social introvert who doesn't like to work with people, why do you run a blog where you socially interact with people? Indeed, not just "people," but strangers.

I'm an introvert, but I'm not anti-social. I prefer to work alone only when it's necessary or more efficient. Likewise, I prefer working with people when it's necessary and more efficient. I recently helped my sister move all of her furniture to a new house, and I was downright annoyed that I couldn't get any other men - including my brother - to work with me. :-)

There's another, more important context in which I prefer to work with people. In my search for finding solutions to the world's big problems, I think it's necessary and invaluable to bounce my ideas off other people. I want my ideas to be tested and challenged by skeptics. In my career as an engineer, I've gone out of my way to ask other engineers to find faults in my designs.

Almost to a fault, I ask people to tell me why I'm wrong about things, even minor things like my opinion about movies. A good example of this is the classic film "Casablanca." This is a beloved film, but I detest it. Whenever I meet someone who says Casablanca is their favorite film, I'll always take the opportunity to explain why I detest it, and I encourage them to tell me why I'm wrong. So far, no one has been able to say why I'm wrong. They'll listen to the reasons I detest it, and, to my surprise, all of them admit that they hadn't noticed the parts I mention. They've all said that I make good points but they need to think about it, or maybe re-watch it, before they can respond. But since it's just a movie, none of them has had enough interest to respond. So I'm still waiting for a fan of Casablanca who is interested enough to respond.

Okay, that's a long-winded way of saying that it's sometimes fun, good and necessary to work with people, even if you're an anti-social introvert. You sometimes need other people to get out of your bubble and keep you honest.

Did I need to write this much to get my points across? No, but you know what they say: It takes more effort and time to edit than it does to write. When something appears to be too long to invest the time to read and digest, professional editors refer to it as TL;DR (too long; didn't read).

Steven Mason said...

Sabine wrote: I am 42 years old and have a temporary position on a contract that will run out next year. I may be many things, but successful I am not.

I want to say something and please don't interpret it in any way negative. Your chosen career is physics and I hope you achieve whatever kind of success that you desire.

That being said, let me say, out of the clear blue, that I would read any novels you would care to write, if they feature physics. I enjoy reading novels that feature science, and I'm not alone. You're a good writer, you're creative, you have obvious artistic urges (e.g. your music), you have a good imagination and a good sense of humor. You're an anti-social introvert who doesn't like to work with people, which isn't a problem for authors. If any of your novels is successful, it could be a source of income.

Let me give a small sample of novels I've read featuring science: A Hole in Texas (particle accelerator), The Chemist (a female government agent who is a chemistry expert), The Martian and Artemis (Mars and the Moon), Chemistry (a partly autobiographical novel about a female chemistry graduate student).

There is also narrative nonfiction that I find very enjoyable. For example, there is How to Bake Pi: An Edible Exploration of the Mathematics of Mathematics, written by a female mathematician. There are tremendously successful books like Hidden Figures and The Immortal Life of Henrietta Lacks. Mary Roach has written a number of successful science-related books such as Gulp, Stiff, Bonk, and Packing for Mars.

I'd love to see more novels and narrative nonfiction that feature physics of any kind, and/or the world of theoretical physics and academia. Novels can include elements of science fiction.

Even if you enjoy writing, it takes a lot of time that you may not have. But if you've got the time and inclination, go for it.

Michael John Sarnowski said...

I think, if there are less citations for women than men, it is probably related to gender discrimination and what the strongest goals of the female researchers are.

Ian Miller said...

One question: how do you deal with "big science", where huge funding has gone to one piece of equipment. An example I know of is the Mars Rovers. NASA encourages/requires anyone with an interest and competence to be part of a team looking that the data, with the net result that there can be up to 50 authors on a paper. That there are more men than women in that area, and they tend to share citations with related work, this might distort such a study.

Unknown said...

dear Sabine,

using fractional counting (namely, a fraction 1/N_aut of each paper is attributed to each author) the total biblio-metric output of collaborations roughly scales on average as the number of authors in the collaboration. So, if fractional counting is used as a figure of merit, the choice of working alone or in a collaboration does not affect the probability of getting hired. Using total citations makes instead convenient to enter in big collaborations, and this creates bad incentives. For example writing two pages to be put in a report with 100 authors, and repeating this strategy, would be more productive than writing research papers. Big collaborations are needed, but counting 6000 papers written by a collaboration of 3000 authors as 3000 x 6000 papers dominates the data-base, at a point that your old plot could not see gender differences.

The statements in your slide 37 sound paradoxical, given that the various gender citation probabilities are not independent (either you cite a men or a women). I understand that you don't plot probabilities, but quantities obtained by divided by the number of papers. I had tried this and similar quantities (dividing by the number of authors, dividing by the citations from multiple-author papers, etc). But all of them are arbitrary choices. There is no way of fully disentangling a hypothetical gender discrimination from a difference in number or in productivity. Then the way to get safe information on discriminations is finding the observable that projects out differences in number or productivity: this is the "gender asymmetry" I computed.

By the way, about a month ago I submitted a scientific paper with details of my analysis to arXiv, and it's kept "on hold". Before talking, I considered the danger of being censored in silence.

best,
Alessandro Strumia

marten said...

Women in physics or women in science?
There are physicists who are not scientists and there are scientists who are not physicists.

naivetheorist said...

bee:

"The reason I do well on this particular index is simply that I am an anti-social introvert who doesn’t like to work with other people. And, evidently, I am too old to be apologetic about this. Since most of my papers are single-authored, I get to collect my citations pretty much undiluted, in contrast to people who prefer to work in groups."

i think the reason may be that you're the author of a popular and highly regarded scientific blog and a best-selling (at least i hope it's best-selling) book. and the citations to your work are part of the 'rich get richer' phenomena (see https://www.nytimes.com/1995/09/24/books/why-the-rich-get-richer.html and Robert H. Frank's book "The Winner Take All Society").

of course, another reason might be that you write important and significant articles that others recognize as such.

best regards,

naive theorist

naivetheorist said...

To Steven Mason:

i'd like to recommend two old and long forgotten novels: C.P. Snow's first novel "The Search" and Eckert's "The Hab Theory" as excellent novels about the actual practice of doing and presenting scientific research.

naive theorist

Sheever said...

You re not anti social, maximum asocial.

Federico Fuentes said...

Hi Sabine,

Very interesting presentation. I have a few questions regarding the proportion that each of the two reasons you found (1. less female single-author papers, 2. fewer citations/paper of female single-authored papers) actually explains the phenomenon of women being cited less.

It seems to me that the first reason on its own may account for most of the difference (assuming the second reason wouldn't exist) if single-authored papers tend to be cited more than the average paper. In other words, if both genders were cited about the same per single-authored papers, but single-authored papers were cited more than the average, then the gender disparity in single-authored papers, which is huge (80% male - 20% female, correct me if I'm wrong), would produce many more citations for males on average. With this in mind it would be interesting (and important) if you provided: 1. Proportion of single-authored papers in relation to all papers, 2. N_cit/N_papers total, N_cit/N_{single-authored-papers} (you only provided the gender-specific data for this measure), N_cit/N_{multiple-authored-papers}.

Naturally, the second reason, fewer citations/paper of female single-authored papers, only compounds the effect, but I disagree that the explanation for it is that "men cite women less", at least from what I can gather from your analysis. More clearly, I think your conclusion in slide 39 "Citations to women's single-authored papers are less than those to men's papers because men cite women less. This has a large impact because there are so many more men than women." is premature and exaggerated. I have several reasons to say this: 1. slide 29 leaves no question regarding the gender-difference in citations/paper of single-authored papers, but the citations themselves, from what I can gather, come from both multi-authored papers and single-authored papers. You only looked at the impact from citations coming from single-authored papers when you made your gendered analysis (as this is what makes sense if you're fishing for gender differences as an explanation to the phenomenon in question, as Strumia did). However, citations of single-authored papers coming from multi-authored papers are, intuitively for me (due to my experience in working and publishing in related fields), the vast majority. Therefore, you should at the very least provide gender-specific data due to citations coming from multi-authored papers: N_{multi-authored-papers-citing-M/F-single-authored-papers}/N_{M/F-single-authored-papers} and N_{F/M-single-authored-papers-citing-M/F-single-authored-papers}/N_{M/F-single-authored-papers} or some related measure. In other words, the (potential gender differences of) citations coming from multi-authored papers could easily dominate the effect of fewer citations/paper of female single-authored papers. 2. In both Strumia's presentation and your own, you try analyze F->M and M->F from single-authored papers, which, as previously mentioned, is a very reduced subset of actual citations and papers (at least in modern times), and it seems to me that the intention is a sort of fishing expedition to ultimately boil down the explanation of the phenomenon to the behavior of a specific gender, implicitly blaming that gender (or alternatively claiming there is "no discrimination" if no difference is observed), even though you actually point this out in slide 40. In Strumia's case, using this reduced subset he claimed no discrimination, and in your case (you didn't claim discrimination exactly, but) you claimed that this was due to the behavior of single-authored male papers citing single-authored female papers less, and said the effect could be large due to the disparity in reason 1 (how large???), and you put this in your conclusion slide, even though the effect is not that clear statistically ("the trend may be spurious"), and as I pointed out it may only explain very little of the overall "fewer citations/paper of female single-authored papers".

Sabine Hossenfelder said...

Dear Alessandro,

While I think it is reasonable that N_cit/N_aut is a better approximation to the "true effort" that goes into a paper than N_cit alone (I have been complaining about the use of N_cit various times, most recently here) it is exceedingly unlikely to actually capture what is going on. Especially in large groups, there will typically be a few people doing most of the work, so assuming a uniform "effort distribution" doesn't make sense. Yet, the information about who actually did the work is not in the existing bibliometric record (though some journals are making an attempt at keeping track of author contributions). Honor authorships are another problem with dividing by N_aut.

" I had tried this and similar quantities (dividing by the number of authors, dividing by the citations from multiple-author papers, etc). But all of them are arbitrary choices. There is no way of fully disentangling a hypothetical gender discrimination from a difference in number or in productivity."

You tried various quantities but decided to only show the one that did not reveal a difference in the citation-behavior from men and women? That seems somewhat odd. Having said this, I agree that they are all somewhat arbitrary. I am afraid there really is no "right" answer. Not like this is particle physics or something ;)

Sorry to hear about your arXiv issues, that's unfortunate. (We were wondering why your paper had not appeared yet.)

Sabine

Sabine Hossenfelder said...

marten,

Our analysis is based on papers in physics, using data-bases that are dominated by high-energy physics in particular.

Sabine Hossenfelder said...

Federico,

The last couple of slides are not directly related to the earlier ones. I used those to explain why I became interested in bibliometric analysis to begin with. The "diversity" refers to a diversity of research styles, as I said explicitly in my talk. Best,

B.

Sabine Hossenfelder said...

Federico,

80:20 is the (approximate) ratio of male/female authors (in our sample). Of course you see a similar ratio in the total number of single-authored papers. The relevant thing to look at in the graph about how often people collaborate is that the curve isn't constant. Women write relatively fewer single-authored papers (compared to, say, papers in 100+ collabs).

Sabine Hossenfelder said...

Federico,

"in your case (you didn't claim discrimination exactly, but) you claimed that this was due to the behavior of single-authored male papers citing single-authored female papers less, and said the effect could be large due to the disparity in reason 1 (how large???), and you put this in your conclusion slide, even though the effect is not that clear statistically ("the trend may be spurious"), and as I pointed out it may only explain very little of the overall "fewer citations/paper of female single-authored papers"."

You misunderstood that. I said that it's not clear if there is a trend in the F->F citations. But those hardly make a difference for the total number of citations because there are so many more men than women. You are right in us not having resolved what happens with multi-authored papers. It's something we want to do before we put out the paper but haven't yet done so, so I am afraid there's not much I can say about it.

Regarding how large is the disparity, the green-dotted lines show the ratios.

Best,

B.

Unknown said...

hi Sabine,

Concerning collaborations, I agree that biblio-metrics cannot tell who is doing the job, so putting a non-informative constant factor 1/N_aut is the best we can do. Still, the finding that the total biblio-metric output of collaborations grows almost linearly with the number of authors suggests that collaborations seem close to the ideal equilibrium limit (namely, authors who sign papers giving little contribution get stopped).

Concerning the "gender asymmetry", I showed only it because it's the only quantity that does not depend on arbitrary assumptions. I agree that it does not fully capture the issue, but the only way to say more is comparing "how things are" with models of "how things should be". My own plot is here:
https://www.dropbox.com/s/rgba7l80b5f8o5y/ProbMF2.pdf?dl=0
It shows the "citation probability" with respect to some "equality" model that arbitrarily assumed rates proportional to the number of authors (rather than to the number of papers; probably there are other differences that now I forgot). Anyhow, I did not include it in my talk because I don't see what one can learn from it; probably the model behind is too crude to say anything.

best,
Alessandro Strumia

JeanTate said...

"looked for differences in citations to papers with male and female authors"

How did you tell male from female, by their names perhaps?

If so, what did you do with names like mine (Jean)?

Sabine Hossenfelder said...

Jean,

Please look at the slides.

JeanTate said...

Thanks Bee.

Ah, I messed up!

Here's what I meant to post, just a minute or so after my last post here:

OK, please delete my as yet unpublished comment ... from the slides you did try to identify gender by name. Do you have a list of "ambiguous" names, one that you could make public?

Also, using the Boston Marathon to check for bias is itself very biased ... for example, not many "Chen"'s in that list I expect, and Chinese (and similar) names are surely very under-represented in Boston Marathon lists cf names on published physics papers (Chen is a common name, and is commonly used for both males and females).

Sabine Hossenfelder said...

Jean,

The distribution by nation will certainly not be the same, but I think it gives us a good impression of how well we do. We implicitly have a list of ambiguous names, but presently no plans to make it public/don't see the point.

Phillip Helbig said...

"The reason I do well on this particular index is simply that I am an anti-social introvert who doesn’t like to work with other people. And, evidently, I am too old to be apologetic about this. Since most of my papers are single-authored, I get to collect my citations pretty much undiluted, in contrast to people who prefer to work in groups."

Let's face it. Of all the mistakes one can make when doing bibliometry (what are criteria for moving from acknowledgements to authorship? must the grant recipient be an author? do people cite people they know? do people cite people they want to know? do people cite papers mentioned in review papers without even reading the former? and so on an so forth), the stupidest is saying that 100 citations on a paper with 40 authors is just as good as 100 citations on a single-author paper.

Giulio said...

Dear Sabine,
I have the impression that the absolute number of citations may not be very representative of the overall value of a scientist, and that perhaps some corrections would need to be applied. I have just two examples
-a citation could be positive ("referring to the work of..."), neutral ("others have formulated different hypothesis"), or even negative ("XYZ wrote this, which is clearly wrong")
-the value of a paper could be considered higher if it continues to be quoted years after it was written. This could mean it was an important paper, there to stay, and not just a "trendy" paper which nobody will remember after some time.

Is there any effort to try to go deeper in the analysis along these or similar directions?

Best regards,
Giulio

Steven Mason said...

naïve theorist wrote: i'd like to recommend two old and long forgotten novels: C.P. Snow's first novel "The Search" and Eckert's "The Hab Theory" as excellent novels about the actual practice of doing and presenting scientific research.

Thanks, I'll take a look at those books. I just now glanced at their plots. "The Search" seems to be an exploration of Judeo/Christian themes, while "The Hab Theory" seems to be about the hypothesis that the Earth's axis rotates by 90 degrees every 6,000 years or so.

Steven Mason said...

Sabine wrote: I am afraid there really is no "right" answer. Not like this is particle physics or something ;)

Would you please just admit that you were joking when you said you don't appreciate American humor? Then I can stop my compulsive efforts to prove that you have a sense of humor.

Stephen Hawking had a sense of humor. It was said that he liked to run his wheelchair over the toes of people he didn't like, making it look "accidental" of course. He rolled over Prince Charles' toes and he regretted not getting the opportunity to roll over Margaret Thatcher's toes. So you see, you're not the only physicist who is anti-social.

For me, the funniest thing Hawking ever did was to prove the theory that no matter what, men can still be jerks. Hawking left his beautiful, devoted wife because, no matter how smart a man is, he's still dumb enough to follow his lustful urges. On the other hand, if any man deserved to have fantasies about nurses, Hawking did. :-)

Steven Mason said...

Phillip wrote: Let's face it. Of all the mistakes one can make when doing bibliometry . . .

There's no doubt that it's tricky business to find meaningful answers, and reasonable people will no doubt disagree on various approaches, methodologies and interpretations.

Let's face it: At the very least, some of us have a nagging and not unreasonable suspicion that women and minorities, as groups, don't get as fair a shake as white men, as a group, overall. How do we observe and measure this, and what do we do about it?

Let's face it, part two: In general, there is still a cultural expectation that women are more on the hook for nurturing and caretaking. Of course I know there have been changes and progress, but it's still there. A while back I seem to recall Sabine saying something about Germany having more "family-friendly" policies than the US. I believe that's the case for most of Europe (and Canada). But I know plenty of Americans who criticize "socialist" family-friendly policies.

I have a brother-in-law who says the biggest shame of his life was that his income wasn't enough to allow my sister to stay at home and be a housewife. The funny thing, to me, is that his shame did not motivate him to help my sister with taking care of the kids and the house. My sister was the classic American story of a woman who worked fulltime at a job and fulltime at home, while the husband watched TV every night. Am I crazy to think that his unwillingness to help with the kids and home was more "shameful" than his wife having a career?

Sabine Hossenfelder said...

Steve,

"Would you please just admit that you were joking when you said you don't appreciate American humor? Then I can stop my compulsive efforts to prove that you have a sense of humor."

If I didn't have a sense of humor I'd long be dead. I don't get American comedy and sit-coms is what I'm saying. I don't laugh about 95% of what The Onion comes up with, that's how terrible I am. (Not that German comedy is much better.) In any case, being funny is hard. It's really hard. If I'm funny, it's by accident, not by design. I think I better stay consistently grumpy.

Phillip Helbig said...

"There's no doubt that it's tricky business to find meaningful answers, and reasonable people will no doubt disagree on various approaches, methodologies and interpretations."

Right.

"Let's face it: At the very least, some of us have a nagging and not unreasonable suspicion that women and minorities, as groups, don't get as fair a shake as white men, as a group, overall. How do we observe and measure this, and what do we do about it?"

No conflict with my reservations about bibliometry.

"Let's face it, part two: In general, there is still a cultural expectation that women are more on the hook for nurturing and caretaking. Of course I know there have been changes and progress, but it's still there."

With regard to this and the point above: yes, but the bigger problem now, at least in some places, are probably knee-jerk reactions because people have the foregone conclusion that old, white, heterosexual dudes (like myself---hey, wait a minute, why am I no longer employed in academia? where did my privilege go?) rule the world and see everything as evidence of this. See the link below for some real bullshit. Recent example: after it was announced that Donna Strickland had won a share of the 2018 Nobel Prize in physics, the internet was full of pundits claiming that Wikipedia editors were hidebound defenders of the patriarchy since she didn't have a Wikipedia page (example, one of many: https://www.businessinsider.de/wikipedia-rejected-donna-strickland-entry-before-nobel-prize-2018-10). As several people then pointed out, the discussion archived at Wikipedia (publicly available) demonstrates that the reasons why there was not yet a page had nothing to do with this and everything to do with violating well documented Wikipedia standards. How many of these people apologized and said "I got it wrong"? Truth is, almost everyone was surprised that she was awarded the prize, and most associate professors (or full professors, for that matter) don't have their own Wikipedia page. As far as I know, no pundit predicted that she would win.

One problem with this is that it detracts from real problems. Those interested and wanting to help then get the impression that all injustices are manufactured. Clutching-at-straws arguments like "only 17 percent of the reference project’s biographies were about women" probably reflects historical facts (which, of course, included prejudice against women), not support for the patriarchy.

I consider myself a feminist by any sensible definition of the term. The problem is that not all definitions are sensible and distancing oneself from kooks should not lead to the accusation that one is some sort of macho.

Surveys and discussion of feminist bullshit:

https://whyevolutionistrue.wordpress.com/2018/12/01/everyday-feminism-promotes-astrology/

https://whyevolutionistrue.wordpress.com/2018/12/01/evolution-denialism-from-the-left/.

Read it (and the documents linked to) and weep.

The problem is that many people on "the left" are so worried that people will think that they are neonazis if they don't say "yes" to all feminist claims, no matter how absurd.

Steven Mason said...

@Sabine

Hooray! The anti-social introvert admits that she'd be dead if she didn't have an "accidental" sense of humor.

You can be a consistently grumpy person and a consistently funny person. They are not mutually exclusive. Some of the things you say literally crack me up, and when I show them to my friends, they laugh too.

In your modified gravity video, my wife noticed that you have long fingernails. I told her you like to scratch the blackboard when you're stuck on a hard problem.

As for American sitcoms, most of them seem to have the same plot: Clueless people saying and doing stupid things. Speaking of clueless and stupid, here's what Aristotle said about humor:

"Comedy, as we have said, is a representation of inferior people, not indeed in the full sense of the word bad, but the laughable is a species of the base or ugly. It consists in some blunder or ugliness that does not cause pain or disaster, an obvious example being the comic mask which is ugly and distorted but not painful."

We can blame the ancient Greeks for bad American comedy.

JeanTate said...

@Sabine: Thanks for the clarification.

Two more observations on your study:

1) How did you treat authors whose "names" are just an initial (or two) plus the "family name"? From memory, there are quite a few papers from authors in the former Soviet Union (and present day Russia, etc) like this. Of course, some people do have a given name that is just one letter! Rare, sure, but presumably they go into your "ambiguous" pile.

2) Did you do any analyses on the "ambiguous" people? As they are composed of some mix of male and female (no aliens!), you'd expect their lines/trends/values/whatever to be somewhere between the "male" and "female" ones (to the extent that those two are statistically distinct). If nothing else, this would be a nice consistency check (e.g. if the ambiguous group is waaay off either the male or female, you likely have some sort of problem - data? assumptions? analyses?)

Steven Mason said...

Phillip wrote: everything to do with violating well documented Wikipedia standards

Interesting. I've edited some Wikipedia articles in order to correct errors, but I'm clueless about the "standards." In fact, I assumed that the standards were quite low, because the errors I corrected were pretty bad. That was more than ten years ago, and maybe they've tightened things up a bit since then.

Phillip wrote: I consider myself a feminist by any sensible definition of the term. The problem is that not all definitions are sensible and distancing oneself from kooks should not lead to the accusation that one is some sort of macho.

We seem to be living in a time when being sensible, reasonable, fair, decent, etc. is seen as radical, ignorant or offensive by multiple sides of every issue. One possible end of humans is a virus that kills everyone, but maybe there's already a virus or some kind of prion that's spreading dementia. :-)

Sabine Hossenfelder said...

Jean,

The gender of authors who never use their first name can't be identified. We haven't looked at the non-identified set but will probably do that for the paper. I don't think there's much to learn from that.

JeanTate said...

Thanks Sabine.

I know you guys are sorta just starting out here, but I feel you do need to address "no first name" cases, perhaps separately from "ambiguous first name" ones. Also, while I agree you are likely to find nothing much of interest when you analyze the "not male or female, as far as we can tell" cases, you do need to do so at some point. How many interesting discoveries have been made when someone looks deeply into "unlikely to learn much" datasets?

David English said...

Identity should have nothing to do with importance of contribution.

For this reason, it may be more important to focus one's efforts on ensuring importance of contribution rather than the identity of the contributor.

Though there may be (probably are) injustice(s) committed, rationalizing the exact opposite is not a sustainable solution.

True sincerity is the manifestation of ideals toward people that have no intention of reciprocation.

Rob van Son (Not a physicist, just an amateur) said...

"Identity should have nothing to do with importance of contribution."

Importance is measured in the number of citations. Citations are given more often to those who are known by the author than to those who are not known to the author. In more colloquial terms: Friends are easier cited than strangers. And citations breed citations, i.e., the Matthew principle.

As a result, the number of citations I my papers collect are also influenced by the number of people that know and trust me. And the people that know and trust me is affected by many things, among others, gender.

Steven Mason said...

David wrote: it may be more important to focus one's efforts on ensuring importance of contribution rather than the identity of the contributor.

David, maybe there should be a study on what we should study. :-)

Sabine is attempting to find out if women in physics get fewer citations than men, and if so, she wants to understand "where the differences come from." Maybe there is evidence of gender bias, maybe not. Since there is a long history of gender bias, don't you think it's reasonable to look at this even if there are other important things to work on?

I still recall my utter amazement at gender bias in symphony orchestras. I had always assumed that orchestras were egalitarian. I was blown away when I learned that orchestras were strongly biased against women, and in most cases they weren't aware of their bias. Orchestras were predominately male, but hardly anyone believed that gender bias was a factor until they experimented with blind auditions. With blind auditions, the number of women selected increased dramatically. As a result, many major orchestras adopted blind auditions. One can find gender problems in unlikely places. Symphony orchestras can apply a relatively easy fix for the problem. If there is a gender problem in physics, steps should be taken to fix it.

Regarding your point about "ensuring the importance of the contribution," if you suspect there's an important problem, then perhaps it should be addressed. But could you please clarify what you think the problem might be?

JeanTate said...

Interesting: today’s blog post is about a paper by J. S. Farnes ... male or female? An arXiv author search turns up Jamie Farnes, Jamie S. Farnes, and J. Farnes, in addition to J. S. Farnes. In your bibliography matching, Bee, how did you deal with these cases? One author? Several different authors? Decide on name alone, or try to match institutions too?

Sabine Hossenfelder said...

Jean,

As I told you, we don't look at individual people. You can check at on website how well the author-id works.

David English said...

@ Steven Mason

"Maybe there is evidence of gender bias, maybe not. Since there is a long history of gender bias, don't you think it's reasonable to look at this even if there are other important things to work on?"

Steven, thank you for your feedback. I have no doubt that there was, is, and always will be bias. Bias takes many forms- humans are susceptible to such frailty of character.

My concern is as follows: what is the end-game of bias studies?

Generally, I think it is safe to say that identities are not as persecuted as before- awareness and eradication of discrimination is something that this era can point to with pride. At some point, the social discussion needs to change to merit, or society risks flipping identity bias 180 degrees.

How is that merit defined? I don't know- I'm only smart enough to know that what I am born with between my legs is less important than my actions.

Steven Mason said...

David wrote: My concern is as follows: what is the end-game of bias studies?

In my lifetime, many things have been done to correct problems with gender and ethnic discrimination and bias. Please clarify what you're concerned about.

David wrote: At some point, the social discussion needs to change to merit

Merit has always been an important part of the discussion about discrimination and bias. For example, in the 19th century Susan B. Anthony said, "Equal pay for equal work."

That being said, we also realized that there are other less obvious aspects of merit, because we don't live in a world where everyone has the same privileges, choices and opportunities.

Alejandro Rivero said...

By the way, the parameter here called "Gender Asymmetry" is one of the variants of "Cohen's Kappa", a characterization of square contingency tables. In network theory, it is better known as assortativity. In statistics it has been always a very disputed parameter because of its dependence on "marginals", ie on the row and column sums. On the other hand, its undirected network version, Scotts' Pi parameter, seems to be less problematic, because the symmetry of the table gives less room to play with normalization. In our university we have been using Scott's to analyze the table of collaborations, instead of the one of citations.