This article is published in the May 2005 issue.

In a More Balanced Computer Science Environment, Similarity is the Difference


Gender differences in computer science tend to dissolve—that is, the spectrum of interests, motivation, and personality types of men and of women becomes more alike than different—as the computing environment becomes more balanced. This finding is emerging from our ongoing studies of the evolving culture of computing at Carnegie Mellon as our undergraduate computer science (CS) environment becomes more balanced in three critical domains: gender, the mix of students and breadth of their interests, and the professional experiences afforded all students.

In contrast, studies of gender and computer science conducted within imbalanced environments, including those carried out at our own institution from 1995-99, point to strong gender differences. A principal finding of the Carnegie Mellon study was that men focused more on programming and women more on applications of computers [Margolis and Fisher 2002]. This led to recommendations for a female friendly/contextual approach to the CS curriculum. We would advise caution. Whether or not it is a good idea to incorporate applications into a particular course should depend on whether it makes sense for the subject matter, for the intellectual and technical skills to be developed, and/or for pedagogical purposes. To do so as a means to promote gender equity may actually help reinforce, even perpetuate, stereotypes. Given the changes we are observing in our evolving student body, we believe this is misguided.

Indeed, we believe that the gender differences observed in the earlier study tell more about the biases in our former admissions criteria (and a limited view of the undergraduate CS major) rather than significant or intrinsic gender differences in potential computer scientists. As we have pointed out [Blum and Frieze 2005], during the latter half of the 1990s, the undergraduate CS major at Carnegie Mellon fed primarily into the booming high-tech industry. The high school computer “geek” had an admissions advantage. Women and men with potential to become computer science leaders, but without long-standing programming experience or commitment, had little chance. The very few women who managed to get in had exceptional academic records.

How Our CS Environment Became More Balanced

When our CS admissions criteria changed in the late 1990s so did our student body. Eliminating prior programming from the admissions criteria (while retaining high standards in mathematics and science) and valuing broader interests—to more closely reflect both our school’s goals as well as rational prerequisites for the major—opened the door for a more diverse student body of women and men.2 Importantly, an outreach program focusing on high school CS teachers resulted in increases in female applicants to our undergraduate program.3 As a result, over the past few years, women have comprised about a third of our undergraduate CS population.

To meet the needs of students entering with varying backgrounds, multiple entry routes were created for the first-year programming sequence. In addition, a one-hour weekly Immigration course for freshmen (inspired by our Immigration course for entering Ph.D. students) had faculty from across the School of Computer Science (SCS) talk about their diverse research interests. These were the only major changes to a curriculum that is still boot camp for CS.

In 1999, the proactive student organization Women@SCS was established to help create an environment in which the new student body could flourish.4 As we have emphasized [Blum and Frieze 2002], “Women@SCS explicitly provides crucial educational and professional experiences generally taken for granted by the majority in the community, but typically not available for the minority participants. Many of these experiences are casual and often happen in social settings. For example, in an undergraduate CS program, male students often have the opportunity to discuss homework with roommates and friends late at night or over meals. Course and job information and recommendations are passed down from upperclassmen, from fraternity files, and from friends. Women students, being in the minority, do not have access to—in fact are often excluded from—these implicit and important advantages. As one proceeds into the professional world, similar phenomena occur.”

Findings from Ongoing Interviews

In the spring of 2002 and again in 2004 we interviewed graduating seniors in CS using a protocol adapted from the Margolis-Fisher studies.5 The 2002 graduating class, like all previous classes, had relatively few women. By their senior year, all three classes following them had significant numbers of women. Thus, we dubbed this 2002 class ‘the class in transition.’ Their unique positioning made us eager to record their views before they disappeared into the world beyond Carnegie Mellon. The 2004 graduating class entered our program after the increases in female enrollment and the creation of Women@SCS.

Already in the class in transition we observed marked changes to the earlier findings of Margolis and Fisher. These changes became even more pronounced in the latter class of 2004. Here we present a glimpse of our findings. A fuller picture of the class in transition can be found in [Blum and Frieze, 2005]; our results on the 2004 class will be covered in an upcoming paper.

Whereas in the earlier Margolis and Fisher study, men tended to view computers as an object of study and women viewed computers as a “tool,” this was one area in which our cohort showed strong gender similarities. We found men and women who enjoyed programming and the “geekier” aspects of computer science, and we found men and women who didn’t. Almost all students saw programming as one part of their interests and the computer as a “tool” for their primary focus, which was applications. The image of “dreaming in code” as the dominant characteristic of male computer science students was clearly being challenged. As one woman from the class of 2004 explained, “It’s always fun to sit down in front of a computer and kind of producing code until something is done and it’s such a good feeling. A lot of the time once I sit down and do programming I find myself living in the cluster for a day without eating or sleeping.” Yet she also doesn’t “want to do it for the rest of my life. I want to combine it with other stuff.”

A man in the transition class reflects on his own transition: “I still find computers to be very interesting. But because the field of computer science has grown as I’ve learned more about it, it’s no longer the computer itself and the programming that is interesting. It’s what can be done with the programs that is now interesting…The computer I see more as a tool now, as opposed to this neat toy.”

Another example of how the view of the field crossed gender lines emerged when students were asked to define computer science. The most common theme to emerge was that computer science meant “problem solving” and a “way of thinking.” As one woman in the 2004 class put it, “I look at computer science as a sort of logic-based way to solve problems.”

Contrary to the findings of the earlier studies, our snapshot of students’ perceptions revealed that the confidence of most women in our cohort had increased by their senior year and had not been “extinguished.” One woman in the transition class made this very clear: “I see myself as one of the best of the best now.” A woman in the 2004 class acknowledges “bumps along the road, but overall I think I’m pretty happy with the way it went.” Another woman in that class gives a roadmap for her increased confidence: “Once you start working on different projects or having more projects under your belt you just feel a little better… Public speaking and having a more professional front is all part of it. And joining a group like Women@SCS really helps because there are plenty of chances to speak, talk and, I think, just growing more as an individual.”

Conclusion

We believe that fundamental misconceptions about computer science, rather than gender differences, are a root cause of gender under-representation as well as the current crisis in the field, i.e., the diminishing interest in computer science on the part of all students [Morris and Lee 2004] [Vegso 2005].

The fundamental misconception, of course, is that computer science equals programming. One of the biggest offenders here is the College Board’s advanced placement (AP) tests in “computer science.” Unlike AP tests in other fields—for example in biology, physics, and economics, where the tests (and the high school AP courses preparing for them) cover deep and even cutting-edge topics in the field—the AP computer science test is almost devoid of intellectual content. Indeed, a perusal of sample tests provided by the College Board (http://apcentral.collegeboard.com/) shows that, for the most part, the tests focus on the idiosyncrasies of the programming language du jour. A student observing the content of the most advanced CS course in high school could hardly be blamed for thinking computer science is programming. With the dot-com bust, why would a bright high school student be excited by a future in programming? AP computer science should be replaced6 by a course exposing the breadth and depth of computer science, perhaps along the lines of Andrew’s Leap,7 a summer program for high school students. Such a course would attract many of the high school students taking advanced mathematics, half of whom today are female.

Very few of the pioneers and current professors of computer science were “hackers.” Many were motivated by their interest in logic and in understanding intelligence and problem-solving. Today, in the twenty-first century, with the increasing ubiquity of computing, women and men with a broader and diverse vision and a deeper perspective are critical for the field and will drive its future. Let’s make sure our educational programs reflect that.

End Notes

  1. Research supported by a grant from the Alfred P. Sloan Foundation. This article is based on “The Evolving Culture of Computing: Similarity Is the Difference,” forthcoming in Frontiers 26:1 (Spring 2005), a special issue on Gender and Information Technology.
  2. These changes were a result of Allan Fisher pointing out to the Admissions Office that “prior programming” was not a predictor for success in the CS major and Raj Reddy requesting Admissions to develop criteria to select for “future visionaries and leaders in CS” [Margolis and Fisher 2002].
  3. Over a period of three years (1997-99), 240 high school AP CS teachers participated in gender gap discussions led by Allan Fisher and Jane Margolis held in Mark Stehlik’s summer workshops [Margolis and Fisher 2002]. The subsequent increases of women in our program can be directly correlated with these workshops.
  4. See, http://www.women.cs.cmu.edu and [Frieze and Blum].
  5. In 2002, 17 (of the 24) women and 16 (of the 129) men in the graduating class were interviewed. In 2004, 32 (of the 52) women and 23 (of the 104) men were interviewed.
  6. At a minimum, the current AP test (and AP course) should be re-named “AP Programming.”
  7. Andrew’s Leap was created at Carnegie Mellon by Merrick Furst and developed by Steven Rudich to interest bright high school students in computer science. The Roboleap component is run by Matt Mason. Through special classes and independent projects, students are exposed to the frontiers of computer science. See: http://www.cs.cmu.edu/~leap/

References

Blum, L. 2001. “Transforming the Culture of Computing at Carnegie Mellon,” Computing Research News 13(5):2, 6, 9.www.cra.org/CRN/issues/0105.pdf
Blum, L and C. Frieze, 2005. “The Evolving Culture of Computing: Similarity is the Difference.” Forthcoming in Frontiers 26:1 (Spring 2005), a special issue on Gender and Information Technology.
Frieze, C. and L. Blum, 2002. “Building an Effective Computer Science Student Organization: The Carnegie Mellon Women@SCS Action Plan.” Inroads, SIGGCSE Bulletin, Special Issue: Women in Computing 34(2):74-78. www.cs.cmu.edu/~cfrieze/paper.html
Margolis, J. and A. Fisher, 2002. Unlocking the Clubhouse, Women in Computing. Cambridge: MIT Press.
Morris, H. H. and P. Lee. 2004. “The Incredibly Shrinking Pipeline is not Just for Women Anymore.” Computing Research News 16(3):20.
Vegso, J. 2005. “CS Bachelor’s Degree Production Grows in 2004; Poised for Decline.” Computing Research News 17(2).

Lenore Blum (lblum [at] cs.cmu.edu) is Women@SCS Faculty Advisor and co-Director of the Sloan- funded Women@IT program. Carol Frieze (cfrieze [at] cs.cmu.edu) is Women@SCS Director and Co-Director of the Sloan-funded Women@IT program. Both are at Carnegie Mellon University.

In a More Balanced Computer Science Environment, Similarity is the Difference