NSF Project Summary Example: Build the One-Page Decision
An original NSF Project Summary example, dependency map, revision test, and compliance checklist grounded in current NSF guidance.
Next step
Choose the next useful decision step first.
Use the guide or checklist that matches this page's intent before you ask for a manuscript-level diagnostic.
Quick answer: Write the NSF Project Summary as a one-page decision document with three explicit parts: Overview, Intellectual Merit, and Broader Impacts. The overview states the problem, objective, and approach; the other sections explain how knowledge advances and which specific societal outcomes the project can produce.
Evidence basis: We reviewed NSF's Preparing Your Proposal, PAPPG proposal-preparation instructions, and proposal checklist on September 8, 2026. NSF controls the live solicitation and policy. The example below is original Manusights guidance, not a funded proposal or agency template.
Official NSF sources define the requirements. Manusights judgment cannot predict an award decision; it tests whether the one-page argument is supported by the full proposal.
Section | Reader decision | Evidence to include | Failure signal |
|---|---|---|---|
Overview | Is there a defined problem and executable project? | Gap, objective, approach, expected outcome, and study boundary | A field-sized problem leads to vague activities |
Intellectual Merit | Will the work advance knowledge? | Unresolved mechanism or theory, decisive test, alternatives, and learning value | Novel technology substitutes for a knowledge advance |
Broader Impacts | Can the work produce specific societal benefit? | Audience, activity, partner or route, output, access, and evaluation | Benefits are generic aspirations detached from the work |
Gather the inputs before drafting
Lock the current solicitation, project objective, central knowledge gap, research plan, success criteria, and broader-impacts activities. If these items are unstable, polishing the summary hides rather than solves the proposal problem.
The summary must agree with the Project Description, budget, facilities, data-management plan, mentoring plan where applicable, and biosketches. A reviewer notices when the summary promises a population, instrument, dataset, or partnership that the rest of the application cannot support.
Original worked example
This fictional project demonstrates the three-part logic without reproducing a funded application or implying that one wording pattern determines review outcome.
Overview
Coastal heat forecasts often miss neighborhood-scale exposure because regional models smooth over shade, building form, and pedestrian movement. This project will test whether a transferable street-scale model combining fixed sensors, mobile transects, and interpretable spatial features improves out-of-sample heat-exposure estimates across three cities. The team will preregister the primary comparison, publish a harmonized dataset and model cards, and evaluate transfer to a city not used for training.
Intellectual Merit
The work tests when environmental exposure models travel across urban form rather than only whether one model fits one city. Controlled ablations will separate the contribution of shade, morphology, mobility, and sensor density. The project will advance knowledge by identifying the conditions under which local calibration is necessary and by quantifying uncertainty when those conditions change.
Broader Impacts
The project will co-design a planning dashboard with municipal heat teams, release documented teaching modules for community-college geospatial courses, and train paid student researchers from partner institutions. The team will evaluate adoption through documented planning uses, course completion, and pre/post competency measures rather than treating public availability as impact by itself.
Why the example works
The overview names the decision, design, comparison, and boundary. Intellectual Merit states what will become knowable even if the favored model fails. Broader Impacts identifies people, activities, outputs, and measures. Each sentence has a counterpart elsewhere in the proposal.
Stress-test the dependencies
Read the summary once as a program officer and once as a reviewer outside the narrow specialty. Circle every noun that requires evidence: dataset, partner, population, instrument, model, outcome, and evaluation. Then verify each against the proposal.
The sharpest test is counterfactual: if the primary result is null, what knowledge still advances? If the broader-impact activity has low uptake, how will the team know and adapt? A summary that cannot answer those questions is selling hoped-for success rather than a research plan.
Failure states and recovery
Symptom | Diagnosis | Recovery |
|---|---|---|
Overview reads like an abstract | It reports a topic but not a fundable action | Add objective, approach, comparison, and expected output |
Intellectual Merit repeats “novel” | The knowledge advance is unspecified | Name the uncertainty and the test that resolves it |
Broader Impacts lists broad benefits | Audience, activity, route, and measurement are missing | Convert each benefit into an implemented and evaluated action |
Summary promises more than the budget | Cross-document dependencies are unresolved | Reconcile scope, personnel, equipment, and timeline |
Jargon blocks a broad scientific reader | The summary assumes narrow field knowledge | Define the one term that carries the logic and remove the rest |
Completion checklist
- Confirm the current solicitation and PAPPG instructions.
- Keep the summary within the live one-page limit.
- Use the required Overview, Intellectual Merit, and Broader Impacts structure.
- State the objective, approach, expected outcome, and boundary.
- Name the knowledge advance and the decisive test.
- Tie broader impacts to activities, audiences, outputs, and evaluation.
- Reconcile every promise with the description, budget, team, and supplementary documents.
- Export the final file and inspect the rendered page before upload.
Use the grant proposal example to test the full proposal spine, the NIH Specific Aims example only for NIH-specific review logic, and the research-paper outline guide when the underlying study narrative is still unstable.
Readiness check
Run the scan while the topic is in front of you.
See score, top issues, and journal-fit signals before you submit.
Submit if / think twice if
Submit the summary when each promise has a matching owner elsewhere in the proposal: an aim, method, budget line, team capability, partner role, or evaluation measure. Think twice if the summary contains a polished benefit that the Project Description never operationalizes, or if Intellectual Merit depends on the technology being successful rather than on what the research will teach.
Manusights tests the summary as a dependency map. We trace the objective into the research plan, the claimed knowledge advance into the comparison and analysis, and each broader-impact outcome into an activity and measurement record. This does not grade the proposal as NSF would; it finds contradictions a reviewer should not have to reconcile.
Frequently asked questions
Use Overview, Intellectual Merit, and Broader Impacts, and verify the current solicitation and PAPPG before submission.
No. NSF describes it as a one-page summary of the proposed activity, knowledge advance, and societal outcomes, not merely an abstract of the proposal.
The standard NSF guidance sets a one-page limit, but the current solicitation and proposal type control, so recheck both before upload.
Name the audience, activity, delivery route or partner, output, access plan, and a way to evaluate whether the intended outcome occurred.
Sources
- 1. Preparing Your Proposal, U.S. National Science Foundation.
- 2. PAPPG proposal-preparation instructions, U.S. National Science Foundation.
- 3. Proposal preparation checklist, U.S. National Science Foundation.
Before you upload
Choose the next useful decision step first.
Move from this article into the next decision-support step. The scan works best once the journal and submission plan are clearer.
Use the scan once the manuscript and target journal are concrete enough to evaluate.
Private API processing. Your manuscript is not used to train models.
Put the guidance to work
Choose the practical next step for your research.
Start with the question in front of you, or use the guided learning path when you want a clearer view of the full publishing process.