Hacker News Reader

Is this inherently bad? That would imply that, for systemic reasons, girls are not "brought up" to enjoy computing. Is there harm in girls having different preferences, even if they are driven by the culture? Aren't most preferences driven by culture? You raise interesting points. I believe we should be motivating girls who show interest, but is there harm if no interest is shown?

> I believe we should be motivating girls who show interest, but is there harm if no interest is shown?

The question then becomes why interest isn't shown in the first place.

You could argue, as the "manifesto" writer did, that there's some kind of innate biological explanation. To me, that's a pretty extraordinary claim and requires extraordinary proof. It's true that there are some differences between the sexes. Women do a bit better on some mental tests than men and vice versa. But how much of that is biology and how much is upbringing? For instance, from an early age, studies show parents make girls say "please" more than they do for boys, and emphasize sharing. So do women fair better on some emotional tests because of some innate biology or because we've emphasized that kind of thing literally since they could talk? It's really really hard to tease these things apart. And it's so hard because those differences are tiny. If we assumed those differences really are 100% biological (again, a big assumption), it might explain why Computer Science would be 55% men, and 45% women. It wouldn't explain the huge disparity we see today.

So if the reason women don't show interest isn't biological, what is it? A further complicating factor to the whole discussion is we're talking about why women aren't in computer science now. It hasn't always been this way. Historically, the field of computing had lots of women. A lot of women went from "human computers" who did math by hand to programming the early mainframes (the movie Hidden Figures shows a time at NASA when this was happening). While a gender gap has always existed, it used to be a lot smaller. Women made up about 1/3 of people with CS degrees in the late 70s and early 80s. But as the STEM gender gap has been closing everywhere else, it's actually gotten wider in CS, and only CS. Why? The "it's biology and girl brains are different" argument can't account for that. Biology didn't change.

What did change was society. Movies like Weird Science and War Games (my favorite) came out. The "cool hacker" entered Pop culture, and the stereotype they settled on was a boy. Girls were a lot less likely to have computers in the home as well. This is an interesting article (http://www.npr.org/sections/money/2014/10/21/357629765/when-...) that talks about how early home computers were marketed as a toy to buy for boys.

It's like we as a society decided that computers fell into the "boy stuff" category with GI Joe and baseball, not the "girl stuff" category like Barbie and softball.

I think it all adds up to a pretty convincing argument that biology has nothing (or very, very little) to do with it.

>You could argue, as the "manifesto" writer did, that there's some kind of innate biological explanation. To me, that's a pretty extraordinary claim and requires extraordinary proof.

I don't understand why people think this is such an extraordinary claim. It would be extraordinary if the sexes were exactly equal. There are very obvious and enormous differences between men and women in numerous areas. Why would we expect them to come out exactly equal?

But if you want stronger evidence, try this: http://sci-hub.cc/10.1111/j.1751-9004.2010.00320.x

>Results show that gender differences in Big Five personality traits are ‘small’ to ‘moderate,’ with the largest differences occurring for agreeableness and neuroticism (respective ds = 0.40 and 0.34; women higher than men). In contrast, gender differences on the people–things dimension of interests are ‘very large’ (d= 1.18), with women more people-oriented and less thing-oriented than men. Gender differences in personality tend to be larger in gender-egalitarian societies than in gender-inegalitarian societies, a finding that contradicts social role theory but is consistent with evolutionary, attributional, and social comparison theories. In contrast, gender differences in interests appear to be consistent across cultures and over time, a finding that suggests possible biologic influences.

The point about differences in personality was brought up in the Google memo and perhaps explains your "please" thing. But look at the "interest in people vs interest in things" statistic. A d value of 1.18 is enormous. That means 93% of men are more "thing oriented" than the average women. And that's assuming there is no noise in the measurement, likely the true value is even higher!

There's also this later on in the document:

>When Del Giudice computed the Mahalanobis D for gender differences in Big Five traits, based on a published data set, he found that although the mean d for gender differences in individual Big Five traits was 0.27 (conventionally considered to be ‘small’), the Mahalanobis D was in contrast 0.84, suggesting a relatively large mean separation of men and women in the multivariate ‘space’ of personality

The problem is two-fold. First, how do you distinguish how much of those results are innate from how much are the result of how we socialize girls? Second, how do you get from "men are more thing oriented" to "men prefer programming."

To give an example: In the U.S., teaching is consider a woman's profession, because it involves "nurturing." But in India, teaching is a man's profession, because it involves training the next generation workforce to participate in the economy. Even if you suppose women are more "nurturing" innately, that doesn't explain any particular demographic distribution among, say, teachers.

Another example is the fact that 60% of accountants are women. Accounting is not only about "things" rather than people, but it's often about completely abstract "things" (e.g. accrued tax credits). Women also are more "agreeable." But within, for example, the legal profession, the cooperative practice of transactional law has a lower representation of women than the conflict-oriented area of litigation.

Within STEM: over 40% of math majors are women, but only about 20% of CS majors. In other words, "social" and "people-oriented" women evidently prefer a field where people work almost completely alone on completely abstract concepts. Meanwhile, Pinterest's engineering team is almost all men.

The point isn't whether biological differences exist. The point is whether the evidence for biological differences explains what we observe in reality. And it really doesn't (not without egregious handwaving, anyway).

>First, how do you distinguish how much of those results are innate from how much are the result of how we socialize girls?

There have been a number of experiments trying to remove all cultural influences from kids and see what toys they prefer. Despite their best efforts, boys still prefer trucks and girls prefer dolls. Why should we expect the differences in interests to disappear with age?

Second, where do you get the statistic that 40% of math majors are women? Just a quick Google search and I found this: https://www.theatlantic.com/science/archive/2016/11/math-wom...

>In mathematics, just 15 percent of tenure-track positions are held by women, one of the lowest percentages among the sciences, along with computer science (18 percent), and engineering (14 percent)

They are even worse than computer science. This is about exactly what I would expect. (The article of course blames sexism. Since they can't consider any other explanation. But they don't present any evidence to support that.)

Even if the differences aren't perfectly predictive of real world outcomes, so what? I'm not claiming they are. All I'm saying is we shouldn't expect the distribution of interest to be a perfect 50:50 split between men and women. If the personality differences between men and women are as significant as these studies show, it would be very surprising if every profession had a perfect gender ratio.

Computer programming in particular is extremely interest driven. Programmers like programming and do it in their free time. Most of them taught themselves to do it before they studied it formally. As opposed to your other examples like accounting. How many people do accounting in their free time just for fun?

This is one of those situations where you have to exercise your logical reasoning skills. Why the hell would there be an innate gendered preference for a toy that represents something that's existed for only about a hundred years? Indeed, a 2013 study shows that both girls and boys prefer dolls at 5 months: http://www.sciencealert.com/boys-prefer-dolls-to-trucks.

And even if there was a gender difference in terms of preferring trucks to dolls, what would that really be measuring? Again, a five month old couldn't possibly have an association between trucks and building things. You could just as easily be measuring the degree to which infants have an affinity for faces. You can't get from that to a preference for/against CS without a whole lot of handwaving. Which is ultimately the big problem. You're pointing to biological differences and saying that must explain observed occupational differences, without providing any theory of causation. It's an argument that doesn't just have little explanatory power, what explanatory power it does have is based on gendered characterizations. (E.g. hypothesizing women prefer teaching because it's a "nurturing" profession, while ignoring societies that don't view it as a nurturing profession at all, where teaching is male dominated).

As to math preferences, you're comparing tenure track professors, while I was referring to undergraduate degree holders (hence "math majors"). Women earn over 40% of math/stat undergraduate degrees: https://www.aps.org/programs/education/statistics/womenmajor....

I agree with all your points here and I'd like to especially expand on this:

> (E.g. hypothesizing women prefer teaching because it's a "nurturing" profession, while ignoring societies that don't view it as a nurturing profession at all, where teaching is male dominated).

If women don't go into CS because society as a whole views it as a thing-oriented discipline and women are more people-oriented than thing-oriented, this opens up the possibility of reframing CS to be more people-oriented, changing its perception in society, thus causing more women to become interested in CS.

The memo had this to say on that topic:

We can make software engineering more people-oriented with pair programming and more collaboration. Unfortunately, there may be limits to how people-oriented certain roles at Google can be and we shouldn't deceive ourselves or students into thinking otherwise (some of our programs to get female students into coding might be doing this).

Which actually sounds pretty reasonable to me.

>a 2013 study shows that both girls and boys prefer dolls at 5 months

Might as well just measure what toys they prefer when they come out of the womb. No shit a five month old has no idea what a truck or a doll is. But by the age of four they will definitely prefer trucks as expected.

>You're pointing to biological differences and saying that must explain observed occupational differences, without providing any theory of causation.

I did provide a theory (things vs people), but it's not really necessary. The original comment suggested that there were no personality or behavioral differences between men and women. This is a ridiculous claim. I have showed there are huge differences between the preferences of men and women.

Why would we expect two different groups with (statistically) different personalities and interests to make exactly the same career choices? I can't necessarily predict which factors are the most relevant, or which traits correlate with which careers and how strongly. Those are specifics. But I would be completely shocked if the correlation between mental and personality traits and career choice is absolutely zero. And if they aren't zero, that will necessarily imply a difference between the genders.

The alternative theory is that "it's all cultural". This theory also makes no specific predictions. No matter what outcome is observed, you can blame it on culture. You can invent a just-so story about how mathematicians are less sexist than programmers or something. It doesn't require or present any evidence. It isn't falsifiable. If one cultural factor is ruled out (e.g. discrimination), you can always make another and shove it in (e.g. "role models".)

If you show evidence that one cultural explanation can't possibly be correct, they just move onto another. E.g. showing that there are the same percentage of female tech workers as female computer science graduates. So they move onto blaming sexism in universities. So you show there are the same percentage of girls interested in computer science in high school. And then the theory moves onto blaming parents for discouraging girls from computers... or something. And I can't think of any possible experiment that could test that and prove it false. So they say "aha, gotcha! It was culture all along!" as if they've proved something.

>E.g. hypothesizing women prefer teaching because it's a "nurturing" profession, while ignoring societies that don't view it as a nurturing profession at all, where teaching is male dominated

The link I posted sort of addressed this. In less egalitarian cultures it's possible differences are more cultural. E.g. Indians straight up discriminate against female teachers and don't even make it an option. Or Indian parents forcing their girls to learn computer science because they see it as a way out of poverty. But as cultures become more egalitarian, gender differences will become more determined by biology. The west generally gives children a lot more freedom to choose their career and doesn't discriminate by sex. So gender preference will become a lot more important in such a society.

While that 2013 study [1] did show that both boys and girls prefer faces to objects at that age, the results seem to support boys having lesser face/object preference than girls. Unfortunately raw data wasn't provided so we can only rely on researcher's conclusions that the difference was not statistically significant for female faces/cars, but was for male faces/stoves, although eyeballing the graphs, the only one where error bars don't overlap is the Real face (female) / Real car. Anyway, every face/object pair has boys showing lesser face preference than girls. I really can't see how this can be cited as disproving innate gender preference differences - at best, it fails to _prove_ them, but still hints at their existence.

[1] https://www.ncbi.nlm.nih.gov/pmc/articles/PMC3766397/

This article [0] discusses and expands upon many of the questions you asked.

0: http://slatestarcodex.com/2017/08/07/contra-grant-on-exagger...

> in India, teaching is a man's profession

Any citation on this? i have seen most teachers being female...

Not a hugely reliable source, but this article suggests that most teachers are women in urban areas:

http://www.dnaindia.com/india/report-teachers-day-women-teac...

> To me, that's a pretty extraordinary claim and requires extraordinary proof.

The extraordinary claim here is that there are less women in tech due to sexism from men and we have yet to see the extraordinary evidence.

> So do women fair better on some emotional tests because of some innate biology or because we've emphasized that kind of thing literally since they could talk?

Evolution probably favoured caring mothers, you can observe this in animals as well.

> And it's so hard because those differences are tiny. If we assumed those differences really are 100% biological (again, a big assumption), it might explain why Computer Science would be 55% men, and 45% women. It wouldn't explain the huge disparity we see today.

http://news.nationalgeographic.com/news/2005/08/0831_050831_...

"Chimps, Humans 96 Percent the Same, Gene Study Finds"

4% doesn't really explain why we are totally different form chimps. Not everything scales linear manner.

> Historically, the field of computing had lots of women.

That's both true and not, computing used to be radically different historically then it is right now.

> Why? The "it's biology and girl brains are different" argument can't account for that. Biology didn't change.

What if CS changed or what you do with your CS degree changed?

> that talks about how early home computers were marketed as a toy to buy for boys.

I can debunk that for you: look at countries where computers weren't marketed and have the still ratios.

> It's like we as a society decided that computers fell into the "boy stuff" category with GI Joe and baseball, not the "girl stuff" category like Barbie and softball.

So all companies ignored half of their potential market? It's possible, however, it's hard to explain though why would every single company would throw away half of their market.

> I think it all adds up to a pretty convincing argument that biology has nothing (or very, very little) to do with it.

https://en.wikipedia.org/wiki/Confirmation_bias

Jurassic Park is one of THE seminal movies of the 90s and the female kid in the show saves the day because she "knows Unix". That was pretty cool. Though it was only a small role, not the lead like in War Games.

> If you're like me, you probably thought of yourself as smart, and had a short list of "smart things that would be cool when I grow up." Astronaut, doctor, computers, stuff like that.

I think it was magazines. Which weren't gendered. I don't think I've seen any male programmers on TV, and I don't have family members who were programmers, and I was interested in computers before I knew that programming was a job.

As far as I can tell, computers were just innately interesting.

Well, there's gendered and then there's gendered.

Some things are explicitly gendered, like GI Joe or Barbie.

Some things are implicitly gendered, like Legos. There's nothing about then that makes them "for boys." But most stores put them in the boys aisle. Lego started making "girl sets" that are all pinky not too long ago just so they could go in the girl section.

As a kid, where were the magazines with computer stuff in the store? Were they next to Tigerbeat? Probably not. It's little stuff like that. That's not the whole story, of course. But what we find interesting is at least partly due to society.

The magazines were all together in a single magazine rack.

I'm not saying there's not gendered influences, I'm saying my personal life story doesn't have positive-reinforcing gendered elements that I can tell.