## Monday, October 15, 2018

### How the feminists proved the exact opposite of what they claimed

By now, the manifesto ludicrously titled Particles for Justice – a neo-Marxist edition of 100 Scientists Against Einstein – has collected over 4,000 signatures of left-wing activists and bullies (some of them are already very far from particle physics, however) who decided to team up and try to scare not only Alessandro Strumia but also the whole scientific community.

The number 4,000 isn't increasing too much anymore. Strumia has over 40,000 HEP authors in his database and because by now, even grad (and some undergrad) students etc. have been forced to subscribe to the pamphlet, you see a great confirmation of my previous estimate that 10% of the HEP community actively agrees with similar far left activities. It's a minority but not quite a negligible one – and a very aggressive one.

(Gender Taliban is also going to have an official arm in hard sciences, MeTooSTEM. Some folks behind this ugly brand are just completing their fundraiser to become an NGO. I urge all bosses in STEM to fire everyone who participates in this activity before she or he writes at least 5 papers and collects at least 100 citations – or you will become a victim of one of their Nazi pogroms soon.)

You, a high-energy physicist, have to parrot our lies, otherwise we will make a hell out of your career, as we're trying to do with Alessandro Strumia's, these aßholes are telling everybody. Despite this backlash, almost no one has gone through Strumia's actual arguments. Sabine Hossenfelder may surprisingly be considered a marginal counterexample. Along with a collaborator, she tried to address at least one graph by Strumia.

One of her texts is Gender Bias in Academia: Case Strumia from October 5th and another one is Yes, Women Still Have a Disadvantage, from October 11th.

Here is the only graph she really tries to look at carefully in the older blog post:

You see how the total number of citations (divided among all the authors of a paper) of the average male (blue) and female (red) researcher depends on the scientific age (since the first paper is published). You see the actual depth of the gap rather clearly.

At the beginning, up to the scientific age of 10 years, male and female researchers collect the citations at a very similar rate. This statement doesn't mean that men and women are equally ready to "master the first decade". You must understand that the "average man" whose career path is depicted by the graph is the average of a much higher number of males than the "average woman". When the number of picked male researchers outnumbers the female counterpart 5-to-1 or 20-to-1, depending on the context, then the apparent success of the sexes may look the same for ten years. But again, women must be vastly underrepresented for that gender-blind outcome to emerge.

But even with that success for 10 years, the gap hasn't disappeared for good. You see that it reappears as the scientific age of the researches keeps on increasing. Once the age is over 30 years or so, the number of citations of a man is twice as high as the average female researcher's number and the ratio seems to explode for "really high scientific ages" such as 50 years: the truly lasting results are really found almost entirely by men. This 2-to-1 advantage for men is on top of their 10-to-1 "overrepresentation" in the field – I must emphasize that these two effects aren't equivalent and we're not double counting.

OK, a person who got into the field and who stays in it purely because of affirmative action and politics and who has never found anything of a lasting value may find this graph by Strumia's to be an inconvenient truth. And, lacking integrity, she may want to actively obfuscate it. How did Ms Sabine Hossenfelder obfuscate this self-evidently strong effect showing that when the dust settles, the average male researcher really contributes more than the average female one?

It's simple. To prove that women do as much work as men and they're just being discriminated against and forced to do much more than their male colleagues, she decided what the conclusion should be and ordered a male researcher, Mr Tobias Mistele, to produce some extra graphs that will create the impression that some actual analysis has been done. The irony couldn't be better here. Mr Tobias Mistele was only expected to compute some "details" – but the conclusions were already decided by the truly important researcher in that team, Ms Sabine Hossenfelder. The conclusion must be that there is still a discrimination against women.

What did poor Mr Tobias Mistele do to produce a graph that goes beyond a slide by Alessandro Strumia? He repeated the analysis but only applied it to a subset of the researchers who are "still active" according to a paper-based criterion. What was the resulting graph?

How can you interpret this graph? Well, the blue and red lines almost exactly coincide. There seems to be a universal, sex-blind career path for those researchers who stay active. But do you know what this agreement implies? As the liberal, evolutionary geneticist, and pianist Kisanet Beyene pointed out, it implies that there's no sex bias in citations. The people who really keep on producing papers have a rather universal evolution of their excitement with time and their impact measured by citations as a function of time. In that subgroup of active researchers, the sex and probably other characteristics become rather irrelevant for the scientific process.

This conclusion about the "lack of discrimination in the citation process" is really unavoidable. You couldn't get this precise agreement if there were some substantial anti-male or anti-female bias in the process of citation. Even a small second order residual effect from some bias would probably be large enough to show a clear difference between the male and female curves. But there's no real difference between the two curves. It means that there's no sex-based discrimination when it comes to the collection of citations.

Instead, the big gap only exists when you study the researchers including the inactive ones. It means that the lower output per researcher is smaller for experienced females relatively to experienced males because a higher number of them becomes inactive. What is the cause of this gap? Is it discrimination? Well, if you wanted to say that the discrimination makes the women inactive, you would basically promote a conspiracy theory that says that they can't send preprints to the arXiv etc. Otherwise whether someone writes a paper is clearly up to her or him!

And Sabine Hossenfelder herself is one of those who constantly complain about the "publish or perish" atmosphere. She knows very well that not to publish is a problem. Despite this fact, despite the pressure to "publish or perish", many women simply become inactive after some years. Many of those stay in the system, anyway, which proves that their lack of activity is tolerated much more than men's. Some of the experienced women leave the field – because they lose the interest, the perceived ability, or the citations and other feedback from the peers that often keeps one excited. (The graph above doesn't allow us to determine the relative size of the groups from the previous two sentences.)

So the conclusion is exactly the opposite than what she claims. The "truly intensive" graphs that could detect some discrimination if it existed show a perfect match between the male and female curves. On the other hand, the graphs that primarily depend on the person's decision or ability to publish new papers (before the papers are rated by anybody else!) show a huge and growing gap. To conclude, there is no discrimination against women in the process but there's a significant gap caused by intrinsic differences between the two sexes that matter well before any discrimination could kick in.

Kisanet Beyene has also pointed out a paper by Williams and Ceci published in PNAS 2015 that shows a discrimination against men in tenure-track faculty hires. (The two authors have written lots of related interesting papers.) Women of the same qualities are twice as likely to be hired than the men. Now, this statement is not independent from the gap discussed in the previous paragraph. Instead, they are different ways of talking about the same anti-male bias. At the decisions about the tenure-track faculty jobs, some women are already becoming inactive but they're being hired, anyway.

You know, these obfuscations aren't academic discussions about small corrections to some statements. The left-wing activists want the Academia to hide a huge effect. The composition of theoretical physicists may be 10-to-1 in favor of men but because at the most relevant timescale of 30 years or so, the per-capita output adds an extra 2-to-1 factor in favor of men, the men-to-women ratio of contributions to this hard science is probably 20-to-1 or 95% from men.

In March 2018, Strumia and Torre have provided us with quite some evidence that the actual contribution by men is probably much higher than 95% and probably higher than 99%, despite their having less than 90% of the jobs. So the work done by a man per dollar may beat the female counterpart by 1,000% or so.

Strumia and Torre have employed a Google-style PageRank (they call it PaperRank) mechanism to find the importance of papers. I proposed something similar in 2007 but now I see that already in 2006, I actually wrote about some actual papers with detailed data that had used a similar approach. So with this knowledge about the old blog posts, I think that I shouldn't think that I was an independent co-inventor of that approach.

At any rate, Strumia and Torre have used some imperfect method involving complex Mathematica functions to assign sexes to authors in a database – about 1/6 (not 1/2 as I previously stated) of the authors end up being "sexually undetermined", however. That's too bad and it makes many of their results debatable. Haven't they omitted some of the best females? Couldn't they have manually found the sexes of all the top 500 authors? If they asked someone who (or a group that) knows lots of people in HEP, that work could be rather straightforward (and if there were errors left, surely we would have some understanding for their appearance).

But if you believe that their focus on the "easy identifiable sex" isn't a problem, their rankings of the authors end up with a much more spectacular "advantage for men". On Page 28 of their paper, we read:
Then, the female authors with highest PaperRank are: M. Gaillard (in 114th position), H. Quinn (220th), R.E Kallosh (370th, who becomes top-ranked female in 78th position restricting to papers written after 2000), L. Randall (444th). HepNames does not include old authors like Noether, the Curie, Mayer.
The first woman is on the 114th position of their list, in the "second hundred", and then you see that there's exactly one in each following group of 100 authors: 114, 220, 370, 444. Among the first 113 top authors, women have zero, so it means that even if you cherry-pick to look at the first 114 authors to make it as good for women as possible, they have less than 1/114 of the contributions (because those are a decreasing function). Now, because a woman is added after a group of 100 men, you may easily estimate that it means that the men have contributed 99% or so.

Women have contributed 1% by this counting but by looking at the numbers of employed people, they may be getting 20% of the money or so (21% is the percentage among experimental physicists at CERN) and we are still generously overlooking the non-personal expenses for various pro-women-in-STEM campaigns and conferences. The output-per-dollar may really be 20 times higher for men than for women!

Women are getting about 20% of physics PhD as well as physics Bc degrees these days and the percentage has apparently stabilized in recent years.

Now, I don't propose the policy that women or 95% of women should be fired or things like that just because they're women and therefore a part of this statistics. But we should stop the ludicrous obfuscation of the facts. The fact that men have contributed a great majority of science – even in recent decades when affirmative action was alive and kicking – is so shockingly obvious that every kid with a healthy brain must immediately see it. Every particle physicist, both male and female, simply must have a healthy brain so he or she must be capable of seeing these elementary facts, too. And a scientist must have the integrity to admit the truth that is clearly seen.

You don't have to talk about it. Most people may decide not to talk about these things because they may be considered sensitive. And some men simply want to work in an environment where women exist because it's natural and pleasant – even if it is not justifiable purely meritocratically. Maybe that kind of diversity even increases their output. And the women may even be assured that they shouldn't be afraid of being weaker in research than some men because it's normal to tolerate it more for women. But you shouldn't actively spread self-evident lies or harass others for saying things that are self-evidently true. Such a behavior is utterly incompatible with the integrity that the scientific process existentially depends upon.

And that's the memo.

On a lighter note. The famous ex-director of the Czech National Gallery Prof Milan Knížák has finally joined the MeToo movement:

I have also been harassed by women throughout my life. Already when I was in the cradle, my grandma and my mom were lustfully enjoying themselves when they were changing my diapers, washing my wee-wee, and sweeping my beautiful buttocks.

In the kindergarten, a five-year-old colleague used to jump at me and I could only defend myself by contracting scarlet fever. At the elementary school, it was not just my female classmates who were touching me and, whenever they could, they were lifting their skirts and luring me into their snares. Female teachers were also inviting me to the teachers' lounge where I had to pretend an irritated stomach to prevent them from raping me.

And it continued in my whole life. I was kicked out of universities because I wasn't sufficiently submissive towards my female instructors and their libidinous desires. Even now, when I am already retired, I am afraid of walking in the park because sturdy female pensioners are threatening me even there. (You must surely know the "Witch of the Petřín Hill" who is wandering through the paths of Petřín and harassing even the puppies.)

Me too. Yes. My whole life has been bringing me lots of injustice, too. Women are dangerous and no man can really ever defend himself from them. We should team up. Maybe some kind of a "Men's War" could fix the problem. Let's not be frightened by women and let's not be intimidated by their high number.