The fact that women, minorities, and persons with disabilities remain significantly underrepresented in CISE-related disciplines diminishes us all in our research and education activities, to say nothing of our personal lives. NSF and CISE have long worked to change this situation, but we believe new and strengthened efforts are essential and we are now focusing our attention on doing that.
There have been some successes in addressing gender and ethnic disparities and, thus, there is reason for optimism. Women, for example, are now reaching parity with men in taking four years of math in high school 1. African Americans are approaching their proportional representation in CS undergraduate degrees awarded 2. Overall, though, progress has been too slow and large gender and ethnic disparities remain. The gap is even apparent at the high school level, where increased participation in math and science courses is not reflected in an increased interest in CS. In 2004, U.S. high school girls made up 56 percent of all Advanced Placement test-takers, but made up only 11 percent of those taking the Computer Science AB exam 3. This gender gap was larger than in any other discipline!
That same year, underrepresented minorities made up just 6 percent of the Computer Science AB test-takers. At the other end of the academic pipeline, underrepresentation persists in the research community. In the 2002-03 Taulbee Survey results, [http://www.cra.org/statistics/survey/03/03.pdf], women earned just 16.5 percent of the awarded Ph.D.s in CS&E, and held only 11.8 percent of the tenure-track faculty positions. If we look only at U.S. citizens and resident aliens represented in these data, we find that African Americans, Hispanics, and Native Americans combined made up only 3.7 percent of the awarded Ph.D.s and 2.3 percent of the faculty positions-far less than their proportionate representation in the population.
This under-participation in CS by large segments of our society represents a loss of opportunity for individuals, a loss of talent in the workforce, and a loss of creativity in shaping the future of technology. Not only is it a basic equity issue, but it threatens our global economic viability as a nation. Information technology continues to drive our economy with projected IT job growth through 2012 far surpassing that of the engineering, physical sciences, life sciences, and mathematical occupations 4. In light of the changing demographics of our population, who will fill these jobs? Who will bring the diverse perspectives needed to design technology aimed at global markets? Who will bring the range of perspectives to the research enterprise that can best drive innovation? The CS community must play a role in addressing these issues.
To encourage and enable even wider action by many, CISE is starting a new initiative aimed at broadening participation (BP) in CS. It represents a significant commitment, adding $30M to CISE’s ongoing funding of diversity efforts over the next three years. Its goal is intentionally inclusive: to have our diverse population fully participating in computing research and education. We hope to increase the participation rates of all segments of our society, including majority and U.S. citizens. Initially, though, the program will focus on the groups that have historically under-participated: women, persons with disabilities, African Americans, Hispanics, and Indigenous Peoples including Native Americans, Hawaiian Natives, Native Alaskans, and Pacific Islanders.
We believe that targeted interventions designed to increase participation in these groups will ultimately benefit all. That has certainly been true in the past as, for example, colleges now routinely have all freshmen join orientation programs originally designed to ease the transition for minorities, and many CS departments send their new faculty to CRA Career Mentoring Workshops that were originally designed to assist women new to the faculty ranks.
Each of the underrepresented groups lacks critical mass in the CS community and each has its own set of challenges within its own cultural context. In fact, the situation is complicated because none of these groups is itself homogeneous. There are significant cultural differences within groups that impact access and interest in computing and information technology.
Additionally, women, persons with disabilities, and minorities encounter barriers throughout the educational pipeline, starting with K-12 and continuing through the Ph.D. and faculty ranks. At each stage, and at the transitions between stages, there are different challenges and needs. This gives us a space for intervention that is best described as the cross-product of the underrepresented groups with the educational stages and transitions. Clearly this is a large space. If we considered the local context and institutions, it would be even larger.
Currently most BP efforts are narrowly focused on single points in this space, but CISE has insufficient resources to fund the entire spectrum of those activities. Hence, the CISE BP Initiative will focus on broad alliances (of academia, K-12 outreach, industry, and community-based organizations) across and within targeted groups to address issues spanning wide regions of the space. By using the word “alliance”—a close association of groups formed to advance common interests—we want to capture the idea that the individual groups retain their identity and can continue to focus efforts on issues and challenges unique to their community, while at the same they can come together to leverage work on common issues.
We took the first steps in the Broadening Participation initiative at a workshop run by CRA on October 20-21, 2004, in Arlington, Virginia. Each of the targeted groups was represented at the workshop, along with several social scientists and members of the majority community. The goals of the workshop were to identify community-specific critical issues, catalyze a larger community engagement in BP efforts, identify intra- and inter-community common ground, and promote the formation of synergistic alliances. The participants identified many areas of common ground where efforts could be leveraged:
- Train faculty in cross-cultural mentoring;
- Provide research experiences for undergraduates;
- Develop bridge programs for under-prepared students;
- Work with K-12 teachers in defining computer science curricula;
- Define computer science in order to override popular misconceptions; and
- Develop shared infrastructure and resources for repositories of effective/promising practices, evaluation and assessment expertise, and dissemination mechanisms.
Alliances can give us the impact we need to address underrepresentation broadly, at a national scale, with programs that are both scalable and sustainable. Good models for alliances already exist with the successful, ongoing NSF AGEP Program [http://www.ehr.nsf.gov/hrd/agep.asp] and the newer CISE-supported National Center for Women in Information Technology [http://www.ncwit.org]. It is hoped that the experiences gained in these efforts will inform the creation of the new BP alliances.
As noted above, NSF can encourage and empower efforts, but the innovation and hard work of actually making something happen rests on the shoulders of every member of the computing community. We believe that most people in our community understand and accept, at least in outline form, the multiple motivations for increasing inclusion in our field. Fewer, but a growing number, also have a firm understanding of the very positive value that such inclusion can bring to all of us. Unfortunately, the number of majority members of our community who are actively engaged in increasing inclusion is still small. We are urgently asking the community to support such activities and, as appropriate, to participate. Of course, a few leaders will be needed to formulate and carry out new activities, and we are especially interested in locating and encouraging those people.
We know that everyone is extremely busy and that the common perception is that any call for action must ultimately be balanced by a reduction in some other activity. We don’t believe that perception is true in this case. For most of us, the action that is most needed is simply paying attention to issues of inclusion as we go about our usual activities of teaching, choosing research assistants, helping to hire new colleagues, and supporting the activities organized and led by others. We hope that the efforts we fund with the BP Initiative will succeed in diversifying our community, but that success will depend on all of us committing to make a difference by acting on a personal level.
Peter A. Freeman is Assistant Director of NSF for CISE. Jan Cuny is an NSF/CISE Program Director and Professor, Computer and Information Science, University of Oregon. The authors want to thank Professor Bryant York for his insight and assistance, both with this paper and with fashioning this strategic thrust by CISE.
- Richard J. Coley, Differences in the Gender Gap: Comparisons Across Racial/Ethnic Groups in Education and Work, Education Testing Service, 2001.
- Women, Minorities, and Persons with Disabilities in Science and Engineering: 2004, National Science Foundation NSF 04-317.
- AP Program Summary Report 2004, College Board. Available through http://www.collegeboard.com/prof.
- John Sargent, “An Overview of Past and Projected Employment Changes in the Professional IT Occupations,” Computing Research News, Vol. 16/No. 3, pp. 1, 21.