Preventing sexism in CSE doctoral programs can increase women’s retention. With funding from the National Center for Women & IT (NCWIT), CRA has been studying women in the CRA_W Graduate Cohort program. This program welcomes women graduate students into the computing community and provides them with role models and a broad range of strategies for success. Analyses have produced some interesting findings about women’s retention in CSE doctoral programs.
 
The data indicate that observing or experiencing sexism plays a key role in doctoral women's departure. When asked to describe any sexism (according to their own definitions) that they observed or experienced in their doctoral programs, the Cohort women identified incidents ranging from differential and demeaning to crude and offensive behaviors by some faculty and other students.

A few women faced “male graduate students who openly express their opinions that the women in the program are more likely to be incompetent than the men.” Other women were subjected to behavior that would qualify as harassment. For example, this woman “switched labs because male students frequently and explicitly discussed women and their sex lives in very unsavory ways.” Even faculty members occasionally contributed to making the environment inhospitable for women. For example, “I had a male faculty member state that attendance is so important that missing class to tend to a sick child is unacceptable. In fact, he stated that women with children should "choose" between a family and an education.” His male students apparently were not expected to make a similar choice.

Among Cohort women working toward a doctoral degree in 2008, 12% witnessed or had been subjected to sexism by the spring of their first year. Close to one-quarter of the more advanced graduate women observed or had experienced sexism. These responses suggest that women in CSE doctoral programs perceive less sexism than reported by women in many other settings. Nevertheless, our analyses showed that even this level of sexism is a serious issue in CSE doctoral programs.

Although not a common experience in CSE doctoral programs, sexism leads some women to think of leaving (TOL) their doctoral programs. By the second year of their program, 60% of the Cohort women in our sample thought of leaving. Most often, these thoughts were motivated by low confidence in their own abilities (49% of those who TOL cited this reason), or to take a job (49% of those who TOL). Only 7% of the women who thought of leaving cited sexism observed or experienced as their reason. Nevertheless, sexism has a strong impact on actual departure, unlike other motivations for thinking of leaving.

Many of those who consider leaving persist nonetheless. Comparing women who TOL and persisted with those who TOL and actually left shows that these two groups had very different motivations regarding only one factor—sexism observed or experienced. Thirty-six percent of Cohort women who TOL and left selected sexism observed or experienced as a reason for their TOL, while only 3% of those who TOL but persisted chose this reason. Calculating the odds ratio of actual departure produces only one statistically significant factor (p<=.001). The odds of actual departure are 10 times greater for women who think of leaving because of sexism than they are for any other reason women identified.

In conclusion, sexism seems far from rampant in CSE, but when women perceive it, sexism is likely to be toxic to their persistence. These findings suggest that acting to minimize women’s experience of sexism in their doctoral program could have a measurable positive effect on women’s retention.

This material is based upon work supported by the National Center for Women and Information Technology, the Computing Research Association, CRA-W, and the National Science Foundation under grant number UVA-0413538. Any opinions, findings, and conclusions or recommendations expressed in this material are those of the authors and do not necessarily reflect the views of the National Science Foundation or other supporting organizations. The authors also acknowledge and appreciate the assistance of University of Virginia doctoral student Zhen Wu for her valuable help with data collection and analyses.