> Being a woman of color who has graduated CS107 is almost like being a unicorn, and that’s pretty f'd up.
The real question is to ask "Why don't more women take CS107?" Only about 20% of the Stanford CS department is female. If I assume that a CS107 class is 100 people (it probably isn't--CS107 is not a beginner class--it has data structures as a prerequisite). 20 are women. Stanford racial data suggest that 7% of it's enrollment are African-American.
That means that you get at best 2 "women of color" per year in CS107.
Even if all of these women pass, you're still looking at 10 out of about 10,000? students, at best.
Her experiences may be completely valid, but this is hardly the fault of CS107 or its teacher.
"At Stanford, computer science is the most popular undergraduate major--220 students declared
CS during the 2012-2013 academic year. The introductory CS106A class is 40% female, but
that drops to 30% in the subsequent CS106B and 20% in CS107. The major is approximately 12
percent female."
It would be interesting to see real survival analysis of Stanford CS majors sticking with it.
That's a very interesting downward progression. By just the 3rd course only half the percentage of women, which implies more than half the women quit (assuming the whole class shrunk as well).
Why?
I guess it could be a culture of exclusion, but for it to work that fast it seems like it would have to be exceedingly toxic.
I wonder if Stanford does exit interviews when people leave a major. I had to have informal talks with my advisers (at a different college) when I changed majors, explaining my reasons. I wonder if that data is collected at all? And if not, why not?
> I guess it could be a culture of exclusion, but for it to work that fast it seems like it would have to be exceedingly toxic.
It's a failure of experience.
How many people can bake a cake? More on this later ....
My first assignment in data structures is simply: "Here is Eclipse, here is a project, here are directions for checking that project in, do it. Now use that to create a second project to print "Hello, World!" exactly."
1/3 of the class CANNOT DO THIS and drop. I'm serious. Following directions exactly appears to be a learned skill that not many learn.
Back to baking a cake. This is the same problem. Cakes have some exact directions occasionally. If you get them wrong, something goes wrong. I cannot give out several of my cake recipes because people accuse me of giving them the wrong recipe when the problem is they don't follow it.
And, yet, these recipes were directly out of women's magazines in the 40's, 50's, and 60's. So, it's not an anti-female bias to this characteristic.
It may be more of a fact that women who can follow directions exactly get pulled off into different areas/fields/hobbies before they reach computers.
It seems obvious to me that not having "enough" women in Stanford's CS program isn't Stanford's fault. There's an interest gap that's in place by the time children are in high school, and the numbers I remember reading suggest that it gets slightly better through college and in the workforce.
The real question is to ask "Why don't more women take CS107?" Only about 20% of the Stanford CS department is female. If I assume that a CS107 class is 100 people (it probably isn't--CS107 is not a beginner class--it has data structures as a prerequisite). 20 are women. Stanford racial data suggest that 7% of it's enrollment are African-American.
That means that you get at best 2 "women of color" per year in CS107.
Even if all of these women pass, you're still looking at 10 out of about 10,000? students, at best.
Her experiences may be completely valid, but this is hardly the fault of CS107 or its teacher.