Help for minorities in CS field

Below is an edited version of the executive summary of the Final Report of the Workshop on Increasing Participation of Minorities in the Computing Disciplines. The workshop was held May 4-7 at the Airlie Center in Virginia. The workshop chair was Bryant York of Northeastern University.

The primary motivation for this workshop was the 1991-92 CRA Taulbee Survey, published in the March 1993 issue of Computing Research News. It indicated that less than 1% of the Ph.D.s in computer science that year were earned by African Americans and less than 2% were earned by Hispanics. African Americans and Hispanics make up about 21% of the US population, according to 1990 Census data. No CS Ph.D.s were reported as earned by Native Americans. The low production of minority CS Ph.D.s is just one indication of a lack of participation of minorities within the computing and information science disciplines. Other data indicate similarly low participation throughout all levels.

As a result, we decided to design a workshop that would examine the root causes and make some recommendations that might improve the situation. The workshop was planned in October 1994 and held in May 1995. Our original goal was to consider issues relating to the participation of African Americans, Hispanics and Native Americans. However, because the number of Native Americans enrolled in CS programs is so small, their academic experience is significantly different from that of other minorities. We acceded to the suggestion that we focus the workshop on issues relating only to African Americans and Hispanics. Although the initial goal of the workshop was to identify those issues that are specific to minority students and institutions and to propose remedies, the real strength of the workshop is that we were able to formulate our conclusions and recommendations in generic terms. Minority students and institutions are part of the fabric of diversity in US society.

The most significant conclusions to come out of the workshop were the following:

• A variety of types of mentoring is essential to the success of students at all levels. The base of mentors must be expanded. Same gender and same ethnicity are not required. Scalability of mentoring is a problem.
• The reward structure of academic institutions must be revised to reflect the importance of teaching and mentoring.
• The nation needs explicit metrics for evaluating teaching and mentoring.
• Despite the current glut of Ph.D.s in computer science, some minority institutions should develop Ph.D. programs in computer science.
• The computer science community needs to acknowledge the continuum of need from basic computer/information literacy to Ph.D.-level research.
• All academic institutions--minority and majority--need to clearly articulate their missions. Different missions have different infrastructure requirements.
• The notion of replication of successful federal programs needs to be more carefully analyzed.
• Better coordination among professional organizations, community organizations, industrial organizations and educational institutions with respect to the overall development of children is possible through effective use of the Internet.
• New K-12 curriculum in information management that is integrated with traditional mathematics and science is required.

Our conclusions, for the most part, apply to all students and all institutions. If acted upon through specific projects, we feel that all students would benefit and that minority participation would gradually rise to proportionate levels. To expect a precipitous rise in minority participation resulting from an impulse intervention is folly. The bottom line is that mentoring is the key basic activity. It is a human-intensive activity; at this point, unfortunately, we do not know how to scale it well.

Based on these conclusions, a number of recommendations surfaced during the workshop. Some recommendations were very specific, and others were fairly general. Only a few recommendations were directed specifically at NSF. We attempted to phrase the general recommendations so as to allow flexibility in their implementation by agencies and organizations with specific missions.

• Redistribute research/education dollars to encourage academic institutions to evolve their reward structures to reflect an emphasis on quality teaching and mentoring.
• Publish, in hard copy and electronically, explicit assessment criteria and the evaluation process for all federally funded projects.
• Foster mentoring programs in majority institutions at graduate and undergraduate levels.
• Develop partnerships between minority and majority institutions with strong CS programs.
• Encourage industrial partnerships with minority institutions.
• Provide Internet access to minority institutions--both high schools and colleges. Currently the CISE NCRI Connections program and NASA MUSPIN are doing a good job at the college level. These efforts need to be expanded.
• Provide funding for equipping minority institutions with computing facilities that will support a strong CS curriculum.
• Develop new, more human-oriented and intellectually exciting high school computing curricula.
• Continue and expand, if possible, the Young Scholars, Research Assistantships for Minority High School Students and the Research Experiences for Undergraduates programs.
• Develop community-based networks, and train parents as mentors.
• Study a variety of institutions that have already begun to re-emphasize teaching and mentoring, such as the University of Wisconsin, the City University of New York and so on--with respect to the issue of replicability.
• Study the issue of scalability of different forms of mentoring.
• Provide $20 million per year in matching funds for institutions that attract corporate sponsorship for minority students.

The body of the report contains a complete description of the process of participant selection, the panel topics, summaries of the panel discussions, a list of specific project ideas and important but unaddressed questions.

Appendix A of this report contains a list of often-used acronyms. For full and complete detailed recommendations, the reader is referred to the Working Papers. The Working Papers are available as Northeastern University College of Computer Science Technical Reports, TR NU-CCS-95-08.