Many readers of the Journal have participated in ILI projects, and the program's broad impact on undergraduate programs may seem obvious to us. Nevertheless it is useful to have the program's many strengths documented and its few weaknesses identified. Nearly all doctorate-granting colleges and universities that have undergraduate programs in science, mathematics, engineering, and technology (SME&T) have submitted ILI proposals and 86% of them have received at least one grant. Most non-doctoral four-year colleges and universities (75%) have also submitted, and 55% of these have received one or more awards. The fraction of two-year colleges applying is smaller (40%), but the fraction of proposals funded for four-year and two-year colleges is almost identical-roughly one quarter.

Significant impacts of the program at grantee institutions include reconfiguring, expanding, and updating existing laboratory courses and providing hands-on research opportunities. These benefits continue in nearly all of the institutions long after the end of the grant. The hands-on use of instruments affects the same percentage of women students as in the overall population of undergraduates enrolled in SME&T fields and a slightly higher percentage of underrepresented minority students than in the overall population.

Principal investigators report that their projects benefit other faculty by improving the quality and modernity of courses, the range of courses that can be taught, and the morale of the faculty. However, there were no documented impacts of ILI grants on the faculty reward systems at grantee institutions-improved prospects for tenure, promotion, and salary increases. The ILI grants did have tremendous impact as seed money. On average, during the first five years of a grant, more than four times the grant amount was generated in internal and external financial investments.

Finally, there appears to be a synergistic effect when an unusually large number of awards (more than ten) goes to the same campus. In such cases there were more comprehensive curricular reforms, and changes in curriculum and instruction spread far beyond just the departments that received the ILI grants.

Dissemination of the innovative educational materials and pedagogy created with ILI support has been extensive, but I was somewhat disappointed that only about a third of grantees published their work in journals and other professional media. I certainly encourage submission of reports and materials generated as a result of ILI projects to JCE for peer review and potential publication. Nevertheless, dissemination activities reached about 100 undergraduate professors per project, and the fraction of those who reported incorporating the ILI materials and information into their own courses was generally higher than 50%.

The evaluation identified four areas in which the program had fallen short of its goals: participation by two-year institutions; research-related impacts among undergraduate students and faculty; development of equipment-sharing consortia; and community outreach. If you are developing an ILI proposal, devoting some creative thought to these areas would be a good idea. For example, one could envision an equipment-sharing consortium involving several two-year and one or two four-year institutions. In addition to sharing equipment, such a group could provide community service by doing appropriate chemical analyses of water or other samples or by working with secondary schools to provide access to instruments.

The team of disciplinary experts that participated in the ILI evaluation came to somewhat different conclusions from those of Westat, Inc. Their outlook was considerably more positive, and they suggested that ILI filled a real and increasing need for modern equipment and instrumentation in SME&T programs. The disciplinary experts thought that there was more impact on outreach and undergraduate research than did Westat, and they recommended against focusing the program more narrowly on ILI-rich institutions, despite the apparent synergistic effect. They did suggest that NSF should work to increase the number of proposals from two-year, small, and minority institutions, to increase the number of women and minority principal investigators, and to encourage PIs to disseminate the results of ILI projects more widely.

The ILI program clearly exemplifies NSF's goal of providing catalytic support for science education. The 4074 grants to 1185 different institutions totaling $158.6 million over the first decade have had far more impact than these numbers might indicate. Given the synergistic effect of multiple awards in the same institution, an even higher level of funding would clearly bring even greater benefits. NSF is to be congratulated on a program well conceived and implemented, and the Congress is encouraged to provide much greater support for this and other highly effective NSF programs.