Wednesday assorted links


1: Selection bias. If a woman is able to overcome all the various soft and hard hurdles to becoming a skilled contributor to open source projects, she must have extra dedication, aptitude, etc - so is therefore better than average.

The female coders I know, the ones who would contribute to open source projects - and I only know a few - are not people I would like to mess with. Technically or otherwise.

That applies to males as well. The level of technical proficiency and motivation is abnormal, and very rare.

This is part of the horrifying fact that most software is written by men. A terrible, awful, unjust and obviously sexist situation that needs correcting.

So sit down and write some damn software then. But that is hard, so these uncouth men must be managed to meet our expectations.

Yes, there are barriers for men, but those for women are higher. When I say "barriers" construe it how you like, from institutional sexism to brain chemistry to women having better / more interesting opportunities.

"When I say “barriers” construe it how you like" including not barriers at all.

If you think they do not Nathan, it is up to you to show they do not.

Every study of STEM hiring and Computer Science hiring in particular shows a bias towards women.

"1: Selection bias. If a woman is able to overcome all the various soft and hard hurdles to becoming a skilled contributor to open source projects, she must have extra dedication, aptitude, etc – so is therefore better than average."

Yes, it all makes perfect sense if you just start by assuming what you want to prove.

How familiar are you with open source and free software? Have you contributed, gotten to know the people doing it, why they do it etc.?

My priors are that I know these things partly from my own experience and what I have seen. In that field as well as what I do for a living. There are women in both, but they are few in number. Vanishingly few in number.

It didn't arise because a bunch of women with free time decided they wanted to spend hours staring at a screen thinking in abstract terms on how to structure something very complicated, all the while being reminded constantly how frail the human mind is and how many stupid mistakes you can make during a day. And have those mistakes cause enormous amounts of time to be wasted by others doing the same thing. Which oddly enough elicit rather vigorous language at times.

It arose because a bunch of men did it, and continue to do it.

This is a power play pure and simple. If you want to control what code is written sit down and write the damn stuff. If you can't do it by yourself, get a community of like minded people to assist you, and figure out a way to get it organized and productive.

I guarantee that all the complainers are incapable of starting and maintaining something over years such as we see in the open source and free software products. We are seeing the pioneers in this field pass away, maybe a bit young, but the projects they established are vibrant and active.

So I say to you go write some code, submit it and let smarter people than you judge how valid your opinion is. Otherwise you will be ignored and routed around if you become too much of a nuisance.

Who is complaining? I thought I was just observing. I extensively use OSS and have made small contributions.

> Otherwise you will be ignored and routed around if you become too much of a nuisance.

Not if the SJW's figure out your real identity and attempt to intimidate you by threatening your livelyhood.

Alternatively it could be because in a community which is overwhelmingly homogeneous, being seen to associate with those who are markedly different is a cheap way to stand out from the crowd without going to the trouble of being different yourself. Plus you get bonus virtue XP if the differents are a group which the larger society thinks of as victims. It's a kind of prog version of chivalry.

How could we test this? Perhaps if one analyzed the contributions of another minority in Free Software (which is I assume what you meant by Open Source ;-) Unfortunately we Indians are honorary whites as far as the tech world is concerned and the number of NAMs I know of is even smaller than the number of women.

The reactions to the Github study have been pretty funny. I suspect that if you took the exact same data but labeled it as 'impact of daily sexism in tech now below 2%' it would be hailed as a sign of progress. If asked to estimate a broadly defined sexism penalty, I'd have guessed something higher.

Really the study doesn't show any sexism though

Very Interesting.

>Is sexism rampant on GitHub?

Finally, an easy one I can answer: from this study, which isn't even peer reviewed, the answer is "we have no idea".

Given the data they gathered the correct summary of the study is : "from our data gathering there is none."

But, if they had said that they would have been hounded out of their positions by the left. So they simply changed the summary and went on to live happy lives.

The left, the worst people in the world.

Also, does everything in public life have to be so tedious?

It's not peer-reviewed. I know you link to non-peer-reviewed undergrad studies, but you could give warning.

I wonder what motivated Kareem to start break up and sell his collection. Or at least a sizeable part of it it seems.

Many collectors of any ilk have a problem with letting go. Because their collection also represent a part of their own life's history. It's hard to let go. And when they do it's usually too late.

Kareem is a step ahead. Kudos.

My father in law worked for the UN and spent a lot of time in Afghanistan. I was helping him move some stuff around a few months ago and this included almost 100 large Afghan rugs which he sourced from all over the country.

They're rolled up, sealed and stored in a dehumidified room. There isn't a single one on the floor anywhere.

He will be moving in with us at some point. I will be forwarding him the link to the Kareem story.

Hang one on a wall, and put one or two down on a floor. I bet there are some stunning ones in his collection.

They're beautiful and the stories behind each purchase are great too.

Yes, use them. I do. 2 on the walls the rest on the floor. I don't have 100.

Whatever collegiality Mozart and Salieri were known to enjoy in their day, it was Alexander Pushkin who wrote "Mozart and Salieri" in 1830, the earliest (?) fictional (?) account of the rivalry picked up by subsequent chroniclers.

I've deliberately skipped reading the articles about the friendship between John Paul II and Prof. Tymieniecka because it was clear from the first reports that they had a deep friendship, and that's pretty much the beginning and end of the story. So I didn't read the NY Times article and didn't know about the Houthakker connection. Which is interesting but still the whole story is much ado about nothing, unless one is a biographer of Pope John Paul II.

"What we found shocked us. Women had an overall 2-to-1 advantage in being ranked first for the job in all fields studied. This preference for women was expressed equally by male and female faculty members, with the single exception of male economists, who were gender neutral in their preferences."

From a CNN story on a study by two Cornell researchers.

In the US how do you know whether a job applicant is male or female? There's often no clue in the name e.g. "Tyler".

Even for a Dr. Schuyler Cowen, you look for the pronouns used in letters of recommendation. Not to mention membership in caucuses and associations or fellowships given by organizations like the American Association of University Women. The effort is obviously working.

Most businesses don't hire until after the interview.

Skirts? That might be a hint.

Certainly never before the interview - but you can be pretty sure of the balance of your list of 10 for a preliminary interview round even before you meet them.

There is a fallacy here that is oft displayed in this type of situation where merit is supposed to hold sway.

The correct statistic for hiring processes is the maximal order statistic, not the mean. Alternatively, when the best quantile of a group is chosen, the proper statistic relates to some upper tail of the distribution.

Because there are far more men than women, the tails of the male distribution will be thicker at any given threshold of achievement, even if their mean performance is the same. If you believe that mean performance for men is higher, the more likely the disparity even in the absence of discrimination.

On a large-scale challenge that entails a number of complicated content
initiatives, a lead content strategist could oversee a bigger crew that includes UX practitioners.

Comments for this post are closed