Making Computer Science More Inclusive

Making Computer Science More Inclusive

On Tuesday, October 29, Ashe Dryden visited Stanford’s Department of Computer Science (CS) to deliver an important message about inclusion and the field. “Seventy-seven percent of people in tech are male, and the vast majority are white, middle class, and speak English as their first language,” said Dryden.

Dryden identifies as a white, female, queer computer scientist with a mission to help change the landscape of CS in universities and businesses across the country. Dryden was invited to campus by Amy Nguyen, a junior in CS, who hoped to prompt the Stanford community to think more about why CS suffers from limited diversity here and at other institutions of higher education as well as what that means for industry. Dryden, a programmer herself for twelve years, brought a wealth of professional and personal experience to share.

Much of Dryden’s perspective on diversity in CS is shaped by a theory called intersectional feminism. Intersectionality is a way to describe the interconnectedness of oppressive institutions (be they racist, sexist, classist, homophobic, ableist, etc.). These institutions interact in a way that cannot be examined separately. In her own words, “The way I go through life as a white woman is different than the way a black woman goes through life.” As such, she was careful to garner the input from those around the room who may not have shared her background.

One of Dryden’s main goals is to educate privileged people who hold power to recognize their privilege (an unearned advantage that you get by virtue of the circumstances of your birth and upbringing). When people are sensitive to their own privilege, it is easier to understand others who might be different.

“If you are a white, straight dude who speaks English as your first language, you are probably going to know (and hire) people like you,” she says. This can lead to dangerous uniformity. Dryden consults for multiple start-up companies of Silicon Valley, because “they are the least diverse.” She explains that once a company becomes large enough in the U.S., that company is regulated by much stricter guidelines regarding diversity, and it’s much easier to be sued for infringing upon those guidelines. Most start-ups, however, don’t have the same amount of scrutiny, and recruitment initiatives that prioritize a diverse workforce tend to be overlooked.

The first place Dryden begins is transforming the culture of the start-up. Often a start-up markets their job openings for “fun, young, and casual” workers who are both willing and available to meet the demands of a job that sometimes requires late-night work. “The younger you are, the less likely you are to have kids and a family to take care of. Instead of taking care of your kids, you can keep working and not think as much about what it means for your personal life,” stated Dryden.  But if you ever want to diversify, that approach doesn’t work. And the longer you go without diversity in the workplace, the harder it will be to attract diversity.

Right now, only about 15% of CS graduates in the United States are female, and CS is the only STEM field where that percentage is shrinking instead of growing. Many studies show that at universities across the country far more women are interested in majoring in CS than are graduating in CS. The attrition rate is partially due to the culture of CS departments, many of which “haven’t changed structurally or culturally since they were created some twenty years ago,” says Dryden. The majority of those women lost are weeded out in first year, when departments somewhat deliberately advance students who already come in with basic skills and will be easier to teach and faster to graduate.  

Ashe Dryden, consultant, addressing students in a lecture roomAt Stanford, things are a little different. CS is the most popular major, with some 600 undergraduates currently declared as CS majors. A quick poll of the room estimated that approximately 50% of those enrolled in the booming entry-level course are women. In terms of the numbers, things seem to be improving at Stanford, but participants at the event were quick to point out that culture isn’t always best reflected by a number. For example, much of the late night, all-male culture of industry programming is reflected right here at Stanford. And it starts early on.

One teaching assistant for CS 106a, “Programming Methodology,” which currently has a total enrollment of 609 students, observed that male students in the class have developed networks with other male students who live in neighboring dorms to form coding cliques that simulate the nocturnal coding habits of programmers on the job. This same teaching assistant said that groups seemed to form around gender. Girls remained on the outside and only went to the guys when they needed specific help.

Dryden commented, “This is what male programmers do – drink Red Bull, stay up to 3:00am, and go to hackathons.” Dryden said such activities push away students who have to work, students who have family or other obligations, and students who must rely on a full night’s sleep. To address the issue, Dryden suggested that instructors should create opportunities for students to have access to those same kinds of networks through structured project nights, study groups, and hackathons that take place during the day. Another option would be to have students work on projects during lab time.

Dryden’s informal discussion on how to maximize inclusion within CS led to many useful pedagogical takeaways.

How to make CS classes more inclusive--tips to use now

  • Open the door.  Make sure CS classes are accessible to students who haven’t had the benefit of high school programming exposure or even broadband Internet in the homes their whole lives.
  • Build in quick wins. “We know that women who don’t see early wins blame themselves instead of thinking that ‘maybe this isn’t the best way for me to learn.’ Have short steps built into the course that get students to that next win.”
  • Say it ain't so.  State early and clearly that there is no biological disposition that makes white, male students better programmers than their peers. “There is nothing intrinsic to computer science. A lot of it is cultural.”
  • Provide role models to inspire a diverse group of students. This can be accomplished by hiring diverse faculty and inviting guest lecturers to campus. “Most people in marginalized groups never see somebody ‘make it’ that’s like them,” said Dryden.
  • Own up to your own biases, and help others see theirs. Most - even women - are more likely to doubt the abilities of women than they are to doubt those of men. Moreover, many studies have show that teachers tend to overestimate how frequently they call on female students in class or section. “We affect people on a much larger scale than we think we do,” said Dryden, “so practice apologizing.”
  • Educate students about what their fellow students may be experiencing. Reported Dryden, “We don’t have exposure to the realities of other people’s lives – that is the biggest problem of our industry right now.”
  • Foster collaboration, not competition.  “Imposter syndrome,” the feeling that you must fake being smart enough to make it though an experience because in reality you aren’t really smart enough, tends disproportionately to affect groups who are less represented in CS departments.
  • Don’t separate less well-represented students to try to achieve even distribution. “If you only have two women in the whole section, put them in the same group. It’s a way to advocate on their behalf without isolating them,” said Dryden.
  • Consider an opt-in, female-only section. Reports Dryden, “Studies show that women learn a lot better in groups that are at least 51% female. Men also learn a lot better in groups that are at least 51% female.”
  • Be proactive about making sure students understand. “The worst thing you can do as a professor is to stand in front of a class and ask, ‘Does anybody not know what I’m talking about?’” said Dryden.  Instead, if you think they don’t know something, explain it in a different way. When in doubt, rephrase.
  • Give students study guides so that they can conduct well directed self-learning.
  • Maintain a positive learning environment by accommodating for varying proficiency levels. Saying things like  “This is really basic, but…” or “I know this is really easy, but…” can undermine some students’ confidence and make them even more hesitant to ask a question that might seem too “basic.”
  • Be wary of “benevolent racism,” a way of stereotyping students according to “positive stereotypes” that can have harmful effects on these students. For example, “We have the stereotype that Asian people are smart. We say, ‘You must understand that because you are Asian.’ But this actually forces the students to be compared against other people” and fosters the type of unhealthy competition we should avoid.
  • When a student doesn’t get something, explain it to them.  Don’t do it for them. This has particular ramifications for gendered treatment of students. According to Dryden, some studies have show that high school teachers are more likely to explain a problem to a male student and more likely to do the problem for a female student.

Did you like some of these suggestions? What other suggestions would you like to pose to the Stanford CS community? Please contribute to the comment section below.

Anna Koester Marshall is a doctoral candidate in Iberian and Latin American Cultures.