So, with that, not to mention my nieces, in mind, I thought I’d take an hour (out of finals period!) to hear what Klawe, the dean of the School of Engineering at Princeton, had to say. A lot of it rang true to me, but it was a lot more about organizing and putting together pieces I’d already seen, rather than really telling me anything new.

These are my fleshed-out notes from the talk. The attendance was principally female, maybe 80%/20%, but there was also some significant fraction of the faculty there. Much of the research, Klawe noted, is 4-5 years old.

Computer Science, according to Klawe, once had 30% female majors when it was a new field. What happened?

Myths:

• Computers are a boy toy. (Actually, teenage girls spend more time online than boys.)
• Women can’t do math. (In some Canadian high schools, girls outperform boys in high school math to the degree that there are articles asking what is to be done to help the boys keep up.)
• 24/7 programming is required, and programming is boring, and CS people have no life. (Hmm. We certainly project that image.)
• The computing jobs are all gone now since the dot-com boom burst. (Has nobody heard of Web 2.0?)

These myths may have once been true, but they’re still influencing the way people think.

Truths:

• Boys spend more time playing games.
• Boys and girls like different things in games.
• Games are designed for boys and men.
• Boys who learn to program early do so to create computer games.
• Boys monopolize access to computer at school and at home.

In short, gamers become more computer proficient, and therefore more assertive about computer time.

• Very few high school CS teachers have CS degrees; they end up relying on (male) hackers in class.
• Males in those classes are more prepared and act more confident.
• Assignments in these classes focus on CS, not applications. It was discussed later that women tend to be more interested in outcomes, rather than how it works.
• Software and assignments are often buggy, because we change curricula more often. Female students blame themselves for the bugs (“I can’t make it work!”) rather than the assignment (“It’s a bug in the program.”)
• In North America, gender differences appear early and are sustained into adulthood. Specifically, interest in careers is generally set by ages 10-14.

Expanding on this last point, Klawe showed charts from a study of British Columbia high school students study from 1998-1999. CS and engineering had biggest gap in interest between males and females. Expectation of success followed a similar pattern: more males than females expected to be successful in those fields. This is important because out career choices guided more by personal interests and abilities than anything else. If females think they won’t be as good at CS, they don’t express interest and don’t develop their abilities in it.

Also, once the decision is made to study CS, women experience “impostor syndrome” more often than males. “Impostor syndrome” is described as lack of confidence, and a lack of sense of belonging. It generally comes from feeling different—for example, if you think everyone else you work with is spending their off hours reading slashdot, but you want to be doing something else, like sailing or knitting or walking the dog, you feel different. Men experience impostor syndrome too; I’ve felt it myself, since I didn’t follow the typical teenage-hacker route into CS. But Klawe says women are more likely to take this feeling seriously, and it often causes them to leave courses and careers. (Here she cited Margolis and Fisher.) It occurs at all ages, career points, and levels of achievement.

Have we nailed down the causes yet? Next we started talking about solutions. These fall into three big areas: Increasing girls’ and women’s interest in CS and engineering, increasing their confidence within the fields, and increasing their sense of belonging.

Increasing interest:

• Change the image of CS through media, games, contests, workshops, speakers, programming in math courses. Klawe described her “Oprah project,” trying to get Oprah to run a segment about women whose CS studies changed their lives for the better.
• Emphasize applications
• Combine majors, to emphasize the overlap with nearly everything else. Have computer science majors with (among others,) psychology, biology, art, music, etc. etc. I think this is a really good idea; there need to be more people in (for example) biology who really understand computers, and I think you really need to understand more of what’s possible before you can do something really powerful.
• Include team work, users, communication, volunteer opportunities: we need more “people persons” in CS. (Selfishly, I want to see more women in CS because I think it would lead to a more comfortable environment for me. We tend to be a pretty abrasive lot.)
• Give her a (new) laptop of her own, to counter the boys’ domination of computer time.

Increasing confidence

• How we teach: pairs programming (where students take turns coding and looking over the shoulder “back-seat coding,”) better assignments in labs.
• Unfailing encouragement, positive feedback. This should be obvious.
• Role models and mentors (these don’t actually need to be female, but it helps.)
• Peer cheer-leading groups (self-organizing.)
• Comfy home base—a place where the minority can feel like the majority. I’m not sure I really understood this, but the idea seems to be to create a refuge within the department(s) where women feel they belong.
• Make it OK to cry. (Another one I didn’t completely identify with.) “Tears are frustrating; they have the worst possible outcome,” said Klawe. But women are more likely to cry when frustrated, and programming is frustrating.
• Learning to become strong in an area of weakness. We (I think this applies to men, too,) need to be willing to take on things we’re bad at, and get good at them, rather than only doing things in areas where we expect to be successful.

Increasing sense of belonging

• Achieving critical mass. Apparently 10% representation is enough to make a difference for African Americans.
• Creating environments supportive of personal lives. By this, she meant making it clear that everyone has a life outside CS. It can be as simple as having pictures of your children in your office, or anything that makes it clear that you haven’t signed your whole life over to the department. (I guess my running shirts count?) This reduces impostor syndrome.
• Ensuring inclusive language, images, examples in things like recruiting brochures. You can tell when people are paying attention to this, because you’ll notice women, racial minorities, etc. in their publications. If they’re not paying attention, you’ll see a lot of white males.
• Suppressing jerky behavior. Many sciences have a culture with the idea that if you’re good enough at research, you don’t need to project respect for anyone else. This means that arrogance and disrespect become a sign of success. Obviously, this needs discouraging.
• Treasuring difference of opinion, difference of experience: there is not single right answer to many things, and accepting this idea makes it easier to be the different one.
• Hearing female voices. This is common outside the sciences: woman says something at meeting, is ignored. Five or ten minutes later, man says same thing, and everyone agrees. Dominant ideas shouldn’t always come a from male voice. Klawe said sometimes has a designated male repeater who deliberately brings up points she’s been ignored on, then gives her credit for them, thus pointing out that they’ve ignored her.

None of this is rocket science. But it requires lots of people doing it—it’s about culture change. The good part is, everyone can contribute!

• Encourage
• Provide role models
• Work on the image
• Help high school CS teachers (the JETT project, which I’m not familiar with, was mentioned)
• Use pairs programming
• Make your environment people-friendly
• Hear female voices

Convincing one woman to major, or not drop out, is worthwhile.

Questions

Some raw notes from the questions period at the end. I can’t really dress these up into complete sentences for fear of wiping out the meaning, but I’ve added details where I can remember them.

Q: Women focus on function, men on innards: what’s the outcome in CS that women can focus on?
A: We can find projects with tangible results: for examples, robot control; teaching CS through digital music and media (Georgia Tech) got >60% female enrollment; assignments with a real context (like computational biology work.) Students should be able to do well because they write well and think well, not just program well.

Q: More men moving into the workforce than women; are women dropping out because they’re opting out, choosing family?
A: This is the “faulty NYT studies” question. It’s not true that they’re choosing not to go into the workforce. Women who do go in to the workforce tend to be the happiest ones, more so than men; if we can get them through college and into a job, chances are they’ll be happy there.

Q: Why are girls getting better math grades, but lower test scores?
A: Test scores tend to correllate with confidence; course work is more correllated with knowledge. But this high school problem (grading and evaluating largely on test scores, as well as what’s noted above,) really is a problem, because it means girls are already behind when they’re applying to engineering schools.