Writing a Research Proposal Reviewers Can Evaluate

A research proposal is an argument for work that has not yet been done. Reviewers must decide whether the question matters, whether the proposed approach can answer it and whether the applicants can deliver the work within the available time and resources.

These decisions are rarely based on scientific importance alone. An interesting question can still produce an uncompetitive application if its objectives are diffuse, its methods do not resolve the stated uncertainty or its work plan depends on assumptions that have not been examined.

Begin with the funding call

Read the call before drafting the scientific narrative. Identify the programme’s purpose, eligibility conditions, assessment criteria, required outputs and restrictions on cost or duration. These requirements define the proposal that reviewers have been asked to assess.

A project may be scientifically sound and still sit outside the remit of a particular scheme. Do not rely on a broad thematic resemblance. The application should show a direct connection between the proposed work and the funder’s stated objectives.

Create a short table linking each assessment criterion to the section of the application that addresses it. This prevents important criteria—such as feasibility, research environment, wider benefit or data management—from being left to brief statements near the end.

Define a problem, not a field

The opening section should move quickly from established knowledge to a specific unresolved problem. A broad account of the field consumes space without showing why the proposed work is needed.

A clear rationale establishes four points:

The gap should be supported by evidence rather than declared through phrases such as “poorly understood” or “little is known.” Explain what previous studies could not resolve and why the proposed approach is capable of resolving it.

Write objectives that can be evaluated

Objectives should describe the scientific advances the project will pursue, not simply the activities the team will perform. “Collect samples,” “develop a model” or “analyse the data” are tasks. The objective should state what those tasks are intended to determine.

1. Make each objective specific. Identify the system, relationship or mechanism that will be examined and the uncertainty it addresses.
2. Show how the objectives connect. The sequence should build a coherent programme rather than a collection of related experiments.
3. Avoid complete dependence between objectives. If the first objective fails, the remainder of the project should not become impossible.
4. Define the expected advance. State what will be known, established or made possible when the objective is completed.

Reviewers should be able to identify the purpose of each objective and understand why all of them are needed to answer the central question.

Make the central proposition visible

Most proposals benefit from a concise scientific proposition: an explanation of what the applicants expect to find and why. Depending on the discipline, this may take the form of a hypothesis, conceptual model, design premise or set of research questions.

The proposition should arise from prior evidence and lead directly to the proposed experiments or analyses. It should not appear for the first time halfway through the methodology.

This structure gives the reader a basis for judging whether the planned work can produce an interpretable answer.

Present the approach as a test

A methods section should do more than list techniques. For each objective, explain the rationale for the design, the evidence that will be collected and how the results will be interpreted.

Each work package should address:

• the scientific question or decision being tested;
• the experimental system, dataset or study population;
• the principal comparisons, controls and outcome measures;
• the analytical or statistical approach;
• the result that would support or challenge the proposition;
• technical and interpretative risks;
• alternative approaches if a critical assumption is not met; and
• the output or milestone marking completion.

Methodological detail should be proportionate to risk. Routine procedures may require little explanation. A new assay, difficult recruitment strategy, unvalidated model or complex analytical pipeline requires more evidence and a clearer contingency plan.

Treat feasibility as a scientific claim

Statements such as “the team is well placed to undertake the work” carry little weight without evidence. Feasibility should be demonstrated through the design and the resources available to deliver it.

Relevant evidence may include:

• preliminary data supporting the central premise;
• previous use of the proposed methods or analytical framework;
• confirmed access to equipment, facilities, cohorts or datasets;
• realistic recruitment, sample-processing or data-access timelines;
• expertise covering the major scientific and technical components;
• defined responsibilities among collaborators; and
• a work plan that accounts for dependencies and decision points.

Preliminary results need not anticipate the entire project. Their purpose is to test critical assumptions and reduce uncertainty about whether the proposed work can proceed as planned.

Calibrate the claim of novelty

Novelty is most persuasive when it is defined precisely. A proposal may be original because it addresses a previously inaccessible question, combines methods in a productive way, examines an understudied system or changes how an existing problem can be investigated.

Avoid treating the use of a new technology as sufficient evidence of scientific advance. Explain what the technology makes possible and why that capability changes the question that can be answered.

Claims such as “first,” “unique” or “transformative” should be used only when the literature and the proposed work support them. Reviewers are more likely to trust a well-bounded claim than an exaggerated one.

Write for the actual review process

Proposals are often assessed by panels whose members understand the broader field but may not work on the precise system or technique. The application must therefore satisfy both specialist and adjacent readers.

Define technical terms when they first appear. Explain why a particular methodological choice matters rather than assuming its advantage is obvious. Place essential reasoning in the main narrative; reviewers should not have to infer it from references, diagrams or supplementary documents.

Headings should allow a reader to locate the question, objectives, approach, risks and expected outcomes quickly. Dense pages with little visual hierarchy make scientific evaluation unnecessarily difficult.

Write the summary after the proposal

Draft the summary once the objectives and work plan are stable. It should represent the actual proposal, not an earlier version of the idea.

A useful summary contains:

• the problem being addressed;
• the precise gap in current knowledge or capability;
• the project’s central proposition and objectives;
• the main approach;
• the expected scientific advance; and
• the significance of that advance for the relevant field or beneficiaries.

The summary is not a compressed literature review. Its purpose is to give reviewers an accurate map of the proposal before they encounter the detailed case.

Use internal review to test the argument

Ask at least one specialist and one informed non-specialist to read the application. Give them the funder’s assessment criteria and request criticism rather than proofreading.

Useful review questions include:

• What is the central scientific question?
• Which result would represent the principal advance?
• Where does the project face its greatest technical risk?
• Are the objectives achievable within the proposed period?
• Which statement is insufficiently supported?
• What information was needed but difficult to locate?

If readers give different answers, the proposal’s internal structure probably needs revision. Editing individual sentences will not correct an unclear rationale or a disconnected work plan.

Final proposal check

• The application addresses the stated remit and assessment criteria.
• The unresolved problem is specific and supported by current evidence.
• The central proposition is visible and scientifically defensible.
• Each objective addresses part of the central question.
• The proposed methods can produce an interpretable answer.
• Controls, sample sizes and analytical decisions are justified.
• Major risks and alternative approaches are stated.
• The work plan is feasible within the available time and resources.
• The team’s expertise and access to essential resources are demonstrated.
• Claims of novelty and impact remain proportionate to the evidence.
• The summary accurately represents the complete proposal.
• All funder-specific documents, declarations and formatting requirements are complete.