NSF's success in implementing its core
strategies and achieving its goals depends on how well the agency
manages its investment portfolio and augments it through effective
partnerships with others. Our goal is to ensure an optimum investment
of NSF resources in world-class projects across all programmatic
areas. All the key program functions use the mechanisms of planning
and priority setting, project selection based on merit review
by peer evaluation, and performance assessment, which in turn
feeds into further program planning.
The Foundation's annual planning cycle
provides the means to translate its strategic goals into specific
program design and implementation. Beginning in early spring,
NSF undertakes a planning process to examine its ongoing activities,
and to propose changes to the Foundation's portfolio of programs
consistent with the three NSF goals. This process involves the
interaction of NSF staff, the National Science Board (NSB), and
external advisory contacts, and integrates Administration and
Congressional priorities into the Foundation's plans.
NSF's programmatic priorities are influenced
by a number of factors: scientific and engineering opportunity
and importance, technical feasibility, affordability, and balance
with existing programs. These priority-setting factors are assessed
in discussions with advisory committees, the NSB, professional
societies, the National Research Council, workshops, and task
forces such as the blue-ribbon panel on High Performance Computing.
Planning and priority setting occurs
not only at an agency-wide level, but also within and across individual
disciplines and programs. NSF program officers and management
consult widely with the research and education communities on
opportunities and challenges in all of NSF's programmatic areas,
through workshops, advisory committees, site visits, discussions
at professional meetings, reports and conferences. These varied
inputs help establish programmatic priorities and guide the decision-making
process.
NSF also utilizes its Opportunity Fund
(established in FY 1995) to accelerate investment in emerging
and innovative areas. In FY 1998, NSF plans to provide $30 million
through an Opportunity Fund to pursue exceptionally promising
cross-cutting activities. The Opportunity Fund also often provides
the means for NSF to respond rapidly to unforeseen opportunities
in research and education. In previous years, Foundation-wide
priorities such as research on Life in Extreme Environments and
Learning and Intelligent Systems were enhanced through this mechanism.
Project Selection and Management
The Foundation's most critical priority
setting occurs through the process of merit review by peer evaluation.
NSF currently has the resources to fund about one-third, on average,
of the proposals submitted to it each year. In this extremely
competitive atmosphere, NSF relies on the merit review system
in selecting which specific proposals to fund. With the voluntary
assistance of expert peers in the science, mathematics, engineering,
and education communities, NSF evaluates and selects those proposals
which promise to make the most significant contributions. NSF's
low proposal selection rates and the high quality of numerous
proposals mean leaving many very highly meritorious proposals
unfunded. The selection task is complicated by the increasing
complexity of proposals, many of which are interdisciplinary in
nature, require more complex reviews, or require significant management
by NSF program officers. Nevertheless, the merit review process
provides the best available advice on priorities within and across
programs. NSF program officers, informed by expert peer reviews
from their communities, make decisions on the research to be funded,
decisions which help set the future course of their fields.
At the project level, NSF tailors the
use of peer review in assessing the results of its activities
to the nature of the activity. All investigators applying to NSF
for continued funding must include a statement of the results
of their previous work in their new proposal. These results are
taken into consideration in the new funding decision. These proposals
for renewed support compete directly with proposals for new projects.
For large grants such as for centers, regular site visits are
scheduled and rigorous reviews are undertaken at periodic intervals
during the course of the award. NSF also recompetes centers and
facilities on a scheduled basis. For example, the Engineering
Research Centers and some major facilities operation activities
were recently recompeted. Engineering Research Centers, in addition
to maintaining levels of excellence in their research and operations,
had to demonstrate that they would be doing something significantly
different from their previous work in order to win renewed support
from NSF as a center. Those that could not demonstrate this are
now being phased out. Another example of this type of recompetition
is the current transition from the Supercomputer Centers program
to the new Partnerships for Advanced Computational Infrastructure
program.
Performance Assessment
NSF makes use of two general kinds of
performance assessment. One is the evaluation of NSF's managerial
stewardship of public funds. The other is documentation and assessment
of the results and effects of the investments we make.
Evaluating NSF's managerial stewardship
of public funds focuses on the procedures and decisions of the
program staff. For that assessment, NSF relies on Committees of
Visitors (COV), panels of external experts convened to review
the recent technical and managerial operation of specific NSF
programs. COV procedures were first established for FY 1990, refining
the longstanding External Peer Oversight system. The COV
guidelines were revised in January 1991 to place greater responsibility
on NSF's Assistant Directors for establishing topics to be covered
in each review, and for addressing issues which surface in COV
reports.
Each program that awards grants or cooperative
agreements is reviewed on a three-year cycle. Each COV submits
a report of its findings; NSF management provides a written response
to each COV report. COVs provide a rich source of management information
for NSF. Our current plans for responding to the Government Performance
and Results Act (GPRA) will give COVs an expanded role, to address
not only the management of the merit review process, but also
to assess the performance-the results-of the output of that process.
NSF's active development of performance
indicators that tie to the key program functions of Research Project
Support, Research Facilities, and Education and Training are supported
through the R&RA, EHR, and S&E appropriations. Examples
of the kinds of achievements that will be used to assess NSF's
performance under GPRA are included in the highlights of each
Key Program Function and are described in the following sections.
Back to the Key Program Functions page