Hacker News Board Hacker News, browsed like a classic message board
Live data from Hacker News

How CMU Increased the Proportion of Women in Computer Science (2013)

blog.play-i.com

Post reply on HN
Print view View original

11–20 of 48 posts

Re: How CMU Increased the Proportion of Women in Computer Science (2013)

#11

Post by tjr »

gamegoblin wrote: post #7

I think that the biggest one is number 2 (giving more emphasis to CS potential rather than CS skill in the admissions process). Somehow we've made a society in which getting into computer programming at a young age is generally not something young girls are taught to want. Consequently, the huge majority of young programmers (below college age) are boys. I started programming when I was 12. The advantage this has giv…

I started programming when I was about 11. I did it on my own at home, and also at school, where our classroom had an Apple II computer. At school, I was joined by (especially) two others strongly interested in programming, one boy and one girl.

Looking back, I could not identify anything whatsoever that would have made our programming activities especially appealing to either boys or girls, nor especially undesirable to either boys or girls. Really, I had the impression that pretty much everyone else -- boys, girls, and teachers -- thought our interest in programming was kind of strange, and being a boy (or a girl) had nothing to do with it.

An anecdote, of course, and I have no idea how, or if, it fits into the bigger picture.

Re: How CMU Increased the Proportion of Women in Computer Science (2013)

#12

Post by cschmidt »

khawkins wrote: post #8
cschmidt wrote: post #5

Earlier quoted context omitted.

It sounds like the only thing they did with admissions was to not look at previous programming experience as a major factor. I was admitted to a Chemical Engineering college program, despite the fact that I had no previous high school experience in fluid mechanics, thermodynamics, and the rest. No one thought that was strange. Why is CS so different that you have to be a hot-shot programmer before you enter?

CMU is a top tier school. Why should you deny a hot-shot programmer admission for someone who has a passing fancy? Judging "potential" instead of a candidate's past work and experience reeks of affirmative action.

You are totally using the candidate's past work and experience to judge them. I'm sure those that get in have outstanding AP tests in math and science, great test scores, etc. A hot-short programmer would have that experience counted as well. The point is that programming is not the only type of experience they looked at. That's the same as it is for most other technical degrees (like my ChE).

And there isn't any preference based on gender, so this is nothing like affirmative action. It is just saying that bright technical people can learn programming, and don't need to come in as an expert.

Also, computer science is no more about computers than astronomy is about telescopes [1].

[1] http://en.wikiquote.org/wiki/Computer_science

Re: How CMU Increased the Proportion of Women in Computer Science (2013)

#13

Post by yummyfajitas »

cschmidt wrote: post #5
dengnan wrote: post #2

No offence. But I would be more interested in how they increased diversity without lowering the bar.

It sounds like the only thing they did with admissions was to not look at previous programming experience as a major factor. I was admitted to a Chemical Engineering college program, despite the fact that I had no previous high school experience in fluid mechanics, thermodynamics, and the rest. No one thought that was strange. Why is CS so different that you have to be a hot-shot programmer before you enter?

CS, much like art and music, is something that any sufficiently motivated individual will start doing far before they turn 18. (They differ from chemeng in this manner.) Both CMU art and CMU music require a portfolio review or audition.

http://admission.enrollment.cmu.edu/pages/art-required-portf...

http://www.music.cmu.edu/pages/audition-portfolio-review

Re: How CMU Increased the Proportion of Women in Computer Science (2013)

#14

Post by wrthomps »

I'm a CS major at CMU graduating this year and I think I can help elaborate on the "potential vs. current level" thing a bit. Things have changed somewhat since I entered the school, but the ideas are mostly the same.

For my class, there were three intro-level programming courses, each a prereq for the next: introductory programming for those with no experience, introduction to data structures, and fundamentals of algorithm and data structure analysis. A pre-test given over the summer was used to establish which class you were put into. According to the department, roughly one-third of the class ended up in each one. I had no programming experience coming in, so I was put in the first class.

The final course of that intro sequence was also the major fundamentals course for one of the three tracks that the curriculum branched out to after the first intro class, focusing on algorithms and their implementations. The other two tracks focused on systems programming and mathematical interaction (via functional programming).

Every student, regardless of their intro class, typically completes these tracks by the end of their second year or early into their third year, which opens up basically all of the 300- or 400-level specialization courses in the department. Which is why the school doesn't need to focus on people with lots of experience coming in--the structure of the curriculum means that everyone gets to the same level of important fundamentals pretty quickly, and once you've done that you get to branch out and specialize into whatever you want. The more experienced newcomers just have more time to do that.

And anecdotally, the approach has definitely worked. I've got a great job lined up starting this summer that I'm excited to start. I am not a woman so I can't speak to its effectiveness in attracting women specifically, but in my experience the school has done a great job accommodating a wide range of experience levels.

Re: How CMU Increased the Proportion of Women in Computer Science (2013)

#15

Post by gamegoblin »

khawkins wrote: post #8
cschmidt wrote: post #5

Earlier quoted context omitted.

It sounds like the only thing they did with admissions was to not look at previous programming experience as a major factor. I was admitted to a Chemical Engineering college program, despite the fact that I had no previous high school experience in fluid mechanics, thermodynamics, and the rest. No one thought that was strange. Why is CS so different that you have to be a hot-shot programmer before you enter?

CMU is a top tier school. Why should you deny a hot-shot programmer admission for someone who has a passing fancy? Judging "potential" instead of a candidate's past work and experience reeks of affirmative action.

They aren't giving them negative weight, they are just looking forward.

I know plenty of guys who went into uni with a bit of java experience (AP CS in highschool or something similar). So yeah, they know your basic control flow and OOP. But when it comes to how they will perform in N years, it means nothing. Any intelligent student without experience could (and did!) surpass the "more experienced" programmers by graduation time.

If someone is the next programming wiz of that year, I'm sure they will be viewed as having plenty of potential and get in.

If someone has a year or two of bad Java, they aren't viewing that as any better than someone with a good quantitative background. That's not affirmative action, it's foresight.

Anecdotes, not data, but I know plenty of people who learned Java, C#, or JS in highschool, some even did internships, yet someone who hadn't coded before 2nd year uni soared past their skill level within a year because he was a generally smart guy with good math and logic skills.

So should the people with a bit of experience be assumed to have more potential? No. Should somehow who's been coding their whole life and it's their number 1 hobby be assumed to have more potential? Yes. Not because of their experience, but because of their passion.

Re: How CMU Increased the Proportion of Women in Computer Science (2013)

#17

Post by ivanca »

khawkins wrote: post #8
cschmidt wrote: post #5

Earlier quoted context omitted.

It sounds like the only thing they did with admissions was to not look at previous programming experience as a major factor. I was admitted to a Chemical Engineering college program, despite the fact that I had no previous high school experience in fluid mechanics, thermodynamics, and the rest. No one thought that was strange. Why is CS so different that you have to be a hot-shot programmer before you enter?

CMU is a top tier school. Why should you deny a hot-shot programmer admission for someone who has a passing fancy? Judging "potential" instead of a candidate's past work and experience reeks of affirmative action.

This. When you do something in the name of political correctness (and a completely economically (and risk free) biased one - i.e. almost nobody cares about the lack of female truck drivers) you still have to measure what are you inflicting damage upon.

- Is rejecting someone more fitted for this career path something good?

- Are you assuming everything is a societal result and there are no biological differences between the female/male psyche that makes them look for different hobbies and careers?

- It is know that women care more about socializing than men, and sitting in front of a screen for countless hours is pretty much the opposite of that; even when you work in a team, meetings are usually not something you look up to.

Re: How CMU Increased the Proportion of Women in Computer Science (2013)

#18

Post by gamegoblin »

tjr wrote: post #11
gamegoblin wrote: post #7

I think that the biggest one is number 2 (giving more emphasis to CS potential rather than CS skill in the admissions process). Somehow we've made a society in which getting into computer programming at a young age is generally not something young girls are taught to want. Consequently, the huge majority of young programmers (below college age) are boys. I started programming when I was 12. The advantage this has giv…

I started programming when I was about 11. I did it on my own at home, and also at school, where our classroom had an Apple II computer. At school, I was joined by (especially) two others strongly interested in programming, one boy and one girl. Looking back, I could not identify anything whatsoever that would have made our programming activities especially appealing to either boys or girls, nor especially undesirabl…

It's definitely hard to see. That's part of the problem. But if you look at the number if programmers under age 18, I suspect the gap will be even wider than > 18. Something about society is pushing young women away from it.

Some of my suspicions:

1. The internet is generally more hostile to girls. If you stumble into the wrong chat room or forum, you get "show your tits". When I started programming, I got on the internet for help and met some amazing friends who helped me become a better programmer.

I absolutely hate when gender topics come up on HN because I generally like this website, but gender topics tend to bring the assholes out of the woodwork. I got a female CS student friend of mine reading HN regularly, but when she ran into one of these sorts of threads she was appalled. And rightfully so. And HN isn't nearly as bad as the internet at large.

2. It's tied to video games being male majority. I played computer video games a lot more than any girls I knew as a kid. This exposure to computers (and config files, etc) set the stage for programming.

The video game realm is making a lot of progress in getting girls playing (I think more or consoles than PC, but the general trend is up everywhere), and I've seen this positively influence girls' predispositions towards programming.

3. There is a positive feedback loop in social circles, particularly in formative years (middle school). Kids encourage the status quo. So it's hard for a kid to develop hobbies outside of their clique. I think this relates to the gender problem because girls at that age tend to be even more exclusive than boys (at least in my experience and everyone I've asked).

But those are just suspicions and the real problem could be more fundamental.

Re: How CMU Increased the Proportion of Women in Computer Science (2013)

#19

Post by dengnan »

gamegoblin wrote: post #7

I think that the biggest one is number 2 (giving more emphasis to CS potential rather than CS skill in the admissions process). Somehow we've made a society in which getting into computer programming at a young age is generally not something young girls are taught to want. Consequently, the huge majority of young programmers (below college age) are boys. I started programming when I was 12. The advantage this has giv…

I start programming below college age as well. I also have some friends who learned programming by themselves. The major difference I can tell between those (used to be) young programmers and those who started programming in college is their motivation and passion. Those who started programming as boys/girls usually learned it by themselves and were motivated by curiosity. Knowledge can be learned, experience can be accumulated. But the passion of learning things almost defines one's behaviour. I saw too many people don't want to learn new things because they simply don't like to learn.

Sorry, a little bit OT.

Re: How CMU Increased the Proportion of Women in Computer Science (2013)

#20

Post by sp332 »

ivanca wrote: post #17
khawkins wrote: post #8

Earlier quoted context omitted.

CMU is a top tier school. Why should you deny a hot-shot programmer admission for someone who has a passing fancy? Judging "potential" instead of a candidate's past work and experience reeks of affirmative action.

This. When you do something in the name of political correctness (and a completely economically (and risk free) biased one - i.e. almost nobody cares about the lack of female truck drivers) you still have to measure what are you inflicting damage upon. - Is rejecting someone more fitted for this career path something good? - Are you assuming everything is a societal result and there are no biological differences betw…

But we're not going to be able to measure those things until we try it and find out.

Also, many computer projects are done socially and aimed at socializing. Isolation is an old stereotype.

Post reply on HN