The Effect of Institutional Characteristics on Participation of Women in Computer Science Bachelors Degree Programs Lecia J. Barker Alliance for Technology, Learning, & Society University of Colorado Boulder, CO 80309-0320 Kathy Garvin-Doxas Center for Integrated Plasma Studies University of Colorado Boulder, CO 80309-0320 01.303.492.4643 01.303.735.5578 barkerl@colorado.edu ABSTRACT The poster presents institutional characteristics that can statistically predict the proportion of female computer science bachelors' degrees conferred at institutions. School or college (e.g., engineering) has little predictive power in this model. garvindo@colorado.edu % Female CS BS graduates by institution, 2001 Carnegie classification Nlnst N grads % Female Research: Doctoral Extensive 146 12,662 .21 Research: Doctoral Intensive 91 4,905 .25 Masters College/University I & II 435 .31 15,733 Baccalaureate Colleges 254 4798 .33 Other Institutions 68 2635 .38 Categories and Subject Descriptors A. General Literature 3.2 ^ African Americans, ^ CS Females The higher the proportion of black students at an institution, the higher the proportion of females. In the U.S., females constitute 62% of black undergraduates. At institutions with 80% or more blacks, women received 52% of CS bachelors' degrees in 2000. General Terms Human Factors Keywords Gender, women, IT workforce 1. INTRODUCTION In 1997, Camp [1] showed that computer science (CS) programs in colleges of engineering and prestigious institutions had lower graduation rates of females in CS bachelors' programs than their non-engineering and less prestigious counterparts. The reason, she argued, was that the culture of colleges of engineering are masculine and the majority of engineering students are masculine; thus, a female needs greater confidence for entering CS in engineering than she does to enter CS in other schools or colleges. Camp's data included only institutions that offered a doctoral degree in CS and was based on the 1991-2 and 1992-3 academic years. We wondered 1) what changes have there been in the past ten years? and 2) will a study of all institutions granting BS in CS would yield the same results? 3.3 Private Institutions, ^ CS Females Private schools tend to have higher proportions of CS female graduates than do public schools, at 32% and 26%, respectively. 3.4 Larger Institutions, ^ CS Females Size of institution, measured by total BS graduates, is moderately associated with decreased proportions of females. "College of Engineering Effect"? When entered into a regression model with other factors, the school or college in which a CS program is housed has no significant predictive power. Carnegie classification and size have greater predictive power than college. 4. Conclusion The nature of the institution overall, along with its culture, teaching, reward policies, etc., has more power to predict proportion of female CS BS degrees conferred than does the school or college. While large research programs show smaller female participation, all CS BS-granting institutions need to make programmatic changes to increase the participation of women. 2. RESEARCH METHOD The analysis included 994 institutions granting 41,240 CIS/CS bachelors' degrees in 2001 from U.S. National Center for Education Statistics data [2]. Statistical tests included simple correlation analysis using independent and dummy variables and linear regression. All differences are significant at p < .001. 5.

/lp/association-for-computing-machinery/the-effect-of-institutional-characteristics-on-participation-of-women-b2opLM06Eq