Thus Spake Zuska

“Geek Chic – Computer Science Is The New Sexy”

Posted by Zuska on March 7, 2008
(6)
More »

According to the Chronicle news blog, computer science enrollment is down by half since 2000.

…undergraduate enrollment in computer-science programs had fallen to half of what it was in 2000 (15,958 to 7,915, to be exact).

But according to Inside Higher Ed, the computer science major has rebounded!

For the first time since 2000, the number of newly declared undergraduate majors at doctoral-granting computer science departments is up.

They have a table that shows enrollment for the last 7 years. 2006 does look like the nadir, though it’s a bit soon to tell if the 2007 number is the beginning of an upward swing, or just an anomaly.

Whatever the enrollment trend, we know this:

Martin said, quoting some of her student recruiting ambassadors: “This is geek chic. Our students are getting sexy jobs. Computer science is the new sexy.”

It would be interesting to look at the data for the last decade or so broken out by gender. WEPAN’s site doesn’t include enrollment data for computer science. However, computer engineering data is available from 1996 to 2006.

Though total enrollment in computer engineering grew from 1996 to 2001, the percentage of women remained virtually unchanged. Beginning in 2001, total enrollment declined. Enrollment of females in computer engineering as a percentage went from about 16.0% in 2001 to 11.7% in 2006. When the data is broken down by race, you see a similar trend for women in every group.

Looking at NSF data, total enrollment in all engineering majors rose steadily over the same time period, beginning to decline in 2004. But the percentage of women declined from a peak of about 19.7% in 1998 to 17.2% in 2005.

According to both NSF and the Computing Research Association, the actual number of degrees awarded in computer science increased from 1995 to 2004. That’s where the NSF data ends; the CRA data shows a steep decline beginning in 2003/2004. The percentage of degrees awarded to women declined, however, from 1995 to 2004, from 28.5% to 25.1%. This is the continuation of a trend that began in the mid-1980s, which was the high point for women in computer science.

So much for the slow, steady progress over time theory.

(6)
More »

Comments

  1. #1 kevin
    March 7, 2008

    I’m in compute science, and it feels to me like our field has largely given up on even trying to attract women to the field. The tiny percentage (5%? 10% maybe?) of women in my specialty is probably about the baseline for how low you can possibly get it without (or possibly even with) simply banning women outright.

  2. #2 kezdro
    March 7, 2008

    Looking at enrollment (not graduation), any idea why those numbers would be so low? It’s not going to be the institutions being enrolled in all likelihood (there may or may not be an effect on enrollment by the institution, but if there’s parity in application numbers, I highly doubt it’s THAT much).

    Given what I know of the naive distribution (as in, before specialized training) of the sexes in all relevant areas, I doubt it’s genetic, but what specific environmental conditions could cause this?

    (and, while I think I’m going to check this for myself too, do you know how it works out for mathematics?)

  3. #3 Jane
    March 7, 2008

    Heh. I was intrigued by the joyful tone of the IHE article….and then I saw the numbers. Yes, the number is technically higher, but really, let’s put away the confetti until we see if this super upward trend is really a trend.

    And anecdotally, I’m seeing the number of majors rise, but the number of women remaining constant (or not rising as fast as the number of majors). So again, no cause for celebration there.

    As Kevin points out, some subfields are doing much worse than others in terms of recruiting/retention—it’s definitely not a rare phenomenon.

  4. #4 user
    March 8, 2008

    why would you WANT girls to go into such a boring field full of nerds? They are already dominating biomedical and biological research which is much more interesting, useful, and rewarding.

  5. #5 Becca
    March 8, 2008

    @ user-
    I’m a woman in a biomedical research and I agree it’s very interesting, as well as potentially very useful and rewarding. That said, we *need* more computer scientist expertise in making sense of the awesome power of genomics/proteomics *insert catchphrase ‘omics of your choice here*, as well as a lot of other biological databases.
    Somebody needs to create a training grant for CS/biosciencists that focuses on women and minorities.

  6. #6 absinthe
    March 13, 2008

    As part of my continuing masochistic drive to perform statistical studies of gender trends in science academia, I did a term project for one of my statistics projects last year on gender trends in enrollement and hiring in Physics, Biology, Chemistry, and Computer Science and Math (these studies are masochistic because they always end up with the same conclusion, which never fails to depress me).

    I used NSF data, which doesn’t break down math and CS separately. However, the fraction of people who enroll in math compared to CS is quite small, so the numbers are dominated by CS. I find, as you point out above, the number of people enrolling in CS Bachelor’s degrees dropped. The *fraction* of women enrolling in those degrees gradually but steadily dropped as well. What is interesting is that the fraction of women enrolling in CS doctoral degrees gradually but steadily rose over the same period. A linear fit to both predicts that the fraction of women getting PhD’s and BSc’s in CS will be equal around 2010. Who knows what will happen after that…

    Computer Science is the only science field where the rate of increase of the fraction of female PhD’s per year exceeds the rate of increase of the fraction of BSc’s awarded to females.

    Moving on to employment, the fraction of females in the CS work force vs year of year-of-PhD pretty much exactly matches the distribution of the fraction of PhD’s in CS awarded to women vs year-of-PhD. But, if we look at US born citizens awarded PhD’s and then in the work force, we find CS has the worst unemployment rate for females cf males of all the sciences, and that the disparity is pretty independent of whether or not the people had children. So, US born female computer scientists leak from the pipeline post PhD, and “having babies” isn’t to blame.

    Moving on to salary…female CS PhD’s in the work force have the worst wage disparity compared to their male peers for any of the science fields; of people who received their PhD in the past 10 years, females make around $6k less per year than their male counterparts, and the difference is statistically significant (and I did the analysis weighting the male year-of-PhD to match that of the females to correct for the different seniority patterns between the two groups).

    The wage disparity for recent PhD’s persisted at around $6k even for for childless males and female computer scientists in the work force. For all other science disciplines, the wage disparity was statistically insignificant between childless males and females with recent PhD’s.

    And now I’m depressed just talking about all this. I made it one of my New Year’s resolutions this year to stop doing these friggin’ studies because the conclusions are always depressing, the studies fix nothing, and the resulting frustration is bad for my health…