Skip to content

Structuring a paper

Making the case that your study is worth doing and can be done

A specific question, a method that visibly fits it, and evidence that you can finish the work.

A research proposal argues that a question is worth answering and that you are able to answer it. State the problem and the gap, set out a specific question, describe a method that plainly fits that question, and show the work is feasible in the time and resources available. Reviewers weigh significance and feasibility together.

THE PROCESS

Step by step

Write the question before anything else

Reduce the project to a single question you could put on one line, then check that it is answerable with the resources you will actually have. Proposals built around a topic rather than a question sprawl, because nothing tells you what to leave out. Everything else in the document is downstream of this sentence.

Read the call and build the structure it asks for

Funders and departments publish assessment criteria and required headings, and reviewers score against them directly. Mirror their language in your section titles so a reviewer scoring criterion three can find criterion three. Note page limits, font requirements, and what counts toward the limit before you draft.

Argue for significance in the reviewer’s terms

Explain what will be different if the question is answered, for the field and for whoever else is affected. Ground it in the gap you identified rather than in the general importance of the topic. Panels reject far more proposals for unclear significance than for weak method.

Specify the method concretely

Say what you will do, to whom or what, in what order, and how the data will be analyzed. Name instruments, sampling strategy, and target numbers with the reasoning behind them. A method described in general terms reads as a method not yet decided.

Show the work can be done

Give a timeline with milestones, evidence of access to the site, sample, or archive, and any preliminary work that shows the approach functions. Where equipment, permissions, or collaborators are required, say they are secured or how you will secure them. This is the part reviewers check hardest.

Name the risks and your response to each

Recruitment can fall short, an archive can be closed, a technique can fail to transfer. Identify the two or three most plausible failures and state the contingency for each. A proposal with an explicit risk section reads as more competent, not less confident.

Write the summary last, for a non-specialist

The abstract or lay summary is often the only part every panel member reads, and it is frequently written by a reviewer into their notes. Draft it after the rest, in language a colleague from another department would follow, and make the question and the payoff unmissable.

Have someone outside your subfield read it

Ask them to state your question, your method, and why it matters after one pass. Anything they cannot reconstruct is a section to rewrite, regardless of how clear it seems to you. This single step catches more problems than another week of polishing.

What the reviewer is actually deciding

A proposal reviewer is answering two questions at once: is this worth doing, and can this person do it. Most drafts spend nearly all their effort on the first and treat the second as a formality. That imbalance is the single most common reason a well-conceived project is turned down.

Reviewers also work under time pressure, often reading many proposals in a sitting and scoring against published criteria. Practically, this means your significance claim, your question, and your method should each be findable in seconds — signposted with headings, stated in topic sentences, and not buried in the middle of a paragraph. Assume the document will be read in a few minutes by someone who has already read four proposals that morning.

It also means the proposal is a persuasive document with a specific audience, not a neutral description of a plan. Write it for the panel that exists, using the criteria they have published, rather than for an ideal expert reader who shares all of your assumptions. That is not a reason to overstate the case; it is a reason to be explicit about what you are claiming and why.

Feasibility is the quiet criterion

Feasibility rarely appears as an explicit heading, but it is assessed throughout: whether the sample can be recruited, whether the timeline is credible, whether you have access, whether the analysis is one you can perform. A proposal that promises more than its resources allow gets read as inexperience even when the science is sound. None of these judgments concern the quality of the idea, which is why strong ideas fail here so often.

Concrete detail is what conveys feasibility. “Participants will be recruited from clinical settings” raises the question of which settings and how many; “participants will be recruited from the two outpatient clinics named in the attached letters of support, which together see roughly the required number of eligible patients per month” answers it. The second version is not longer by much and is considerably more convincing.

Scope is the other half. Reviewers see many proposals that would take a decade described as three-year projects. Cutting a project to what fits, and saying plainly that the remaining questions are for later work, strengthens a proposal. An ambitious plan with no evidence of prioritization suggests the work will not be finished.

The parts most proposals contain

The headings vary by funder and by institution, but the underlying components are stable. Follow the required structure exactly where one is given; where it is not, this sequence is a safe default because it moves from why, to what, to how, to when. Where the call sets a page limit for each section, treat it as a statement of how much each part is worth.

Two of these deserve more attention than they usually get. The literature section in a proposal is not a survey — it exists only to establish the gap, and every source in it should be doing that job. The analysis plan is the part reviewers use to judge whether you have thought the project through to the end rather than only to data collection.

  • Title and summary — the question and the payoff, in non-specialist language.
  • Background and gap — what is known, and the specific limitation your study addresses.
  • Aims and research questions or hypotheses — numbered, and reused verbatim later.
  • Method — design, participants or materials, procedure, analysis plan.
  • Ethics and data management — approvals needed, consent, storage, sharing.
  • Timeline and milestones — what happens in which months, with deliverables.
  • Resources and budget — with justification for each significant item.
  • Expected outcomes and dissemination — what will exist at the end and who will see it.

Timeline, budget, and risk

A timeline is a feasibility argument in table form. Break the project into phases with dates, and include the parts people forget: ethics approval, recruitment lead time, transcription, and writing up. Reviewers who have run projects themselves look specifically for these, and their absence is noticeable.

Budget justification is prose, not arithmetic. Each significant line should say why the item is needed and why the amount is right — the participant payment rate that reflects local norms, the equipment that is unavailable in the department, the transcription hours that follow from the number of interviews. Unjustified lines get cut, and a cut line can leave the method unworkable.

The risk section is where you demonstrate judgment. Name the two or three failures most likely to occur, assess how much each would cost the project, and give the mitigation. Acknowledging that recruitment might run slow and stating a second site as backup is far stronger than a proposal that implies nothing will go wrong.

Writing for a mixed panel

Review panels usually include one or two people close to your field and several who are not. The specialists check the method; the non-specialists decide whether the project matters. A proposal that only satisfies the first group tends to score well on rigor and poorly overall.

The practical technique is layering. Open each major section with a sentence any panel member can follow, then add the technical detail underneath it for the specialists. Define the terms specific to your subfield at first use, and keep abbreviations to the few that genuinely save space.

This is also why the summary carries so much weight. Write it so that a reader with no background in your area can say what you will do and why it matters, then check it by asking someone in that position to do exactly that. If they hesitate, the problem is in the summary rather than in their attention.

Doctoral proposals are a related but different genre

A proposal written for admission to a doctoral program or for an upgrade panel is assessed on somewhat different grounds. Alongside significance and feasibility, the readers are judging whether the project fits the department’s expertise and whether you can be supervised in it. Naming potential supervisors and showing you have read their work is not flattery; it is evidence of fit.

Doctoral proposals also tolerate more provisionality. It is expected that the question will evolve, so the document is judged partly on whether you know what would make you change direction. Stating which findings would send the project one way rather than another shows research maturity.

Length varies widely — some departments want two pages, others fifteen — so the ratio between sections matters more than the total. Even in a short proposal, do not let the background crowd out the method; a proposal with four pages of literature and half a page of design is the pattern admissions panels see most often and trust least. Ask the department whether a recent successful proposal can be shared, since these conventions are local.

FAILURE MODES

What goes wrong, and what to do instead

Proposing a topic rather than a question.

Reduce the project to one answerable question and put it early and in its own sentence. Every later section should visibly serve it.

Describing the method in general terms — “interviews will be conducted”.

Give the design, sampling strategy, target numbers, instruments, and analysis plan. Vagueness here is read as a plan not yet made.

Ignoring the assessment criteria published in the call.

Mirror the criteria as section headings and address each one explicitly. Reviewers score against that list, item by item.

Proposing a decade of work in a three-year plan.

Cut to a scope your timeline supports and name the remaining questions as future work. Demonstrated prioritization reads as competence.

Presenting a plan in which nothing can go wrong.

Name the two or three most likely failure points and the contingency for each. A risk section increases confidence rather than undermining it.

Leaving ethics, access, or approvals to be sorted out later.

State which approvals are needed, when you will apply, and how long they typically take — and build that time into the timeline.

Questions writers ask about this

How long should a research proposal be?

Whatever the call specifies, and no longer. Doctoral proposals commonly run from two to fifteen pages depending on the department; grant proposals are usually capped by the funder, often between five and fifteen pages for the case for support. Where no limit is given, aim for the shortest document that fully answers significance, method, and feasibility.

What tense should a research proposal be in?

Future tense for the work you intend to do — “participants will complete two sessions” — with present tense for the current state of knowledge and past tense for what previous studies found. If the proposal later becomes the basis of your thesis or paper, the whole method section has to be converted to past tense, so search for “will” during that revision.

How much detail does the method need?

Enough that a reviewer can judge whether it will answer your question and whether you can carry it out. Name the design, the sampling approach and target number with a justification, the instruments, and the planned analysis. Full protocols and instruments usually go in an appendix if the call allows one.

Do I need preliminary data?

For most doctoral proposals, no. For grant applications in experimental fields it is often decisive, because it converts a claim that the approach works into a demonstration. If you have a pilot, a feasibility study, or even a small dataset, use it as evidence for feasibility rather than as a finding in its own right.

How is a proposal different from a thesis introduction?

A proposal argues in advance that work should be funded or approved, so it is written in future tense and includes timeline, budget, and risk. A thesis introduction reports on work already done, in past and present tense, with no need to justify feasibility. Much of the background can be reused; the argumentative frame cannot.

Read next

Building a literature review that argues instead of lists

Organize by theme, say what the sources mean together, and end on the gap your study addresses.

Describing your method so another researcher could repeat it

Enough detail to reproduce the study, plus a reason for every design decision you made.

From the wide field to your specific question, in about a page

Establish the territory, name the gap, state the question — and stop before it becomes a review.

One workspace
for the whole paper

Draft with structure, cite as you write, and keep a verifiable record of how the document came together. Copyeditor is in beta — apply and we’ll reach out as spots open.

Apply for beta

*Beta opens this season — starting with students and researchers.

The Copyeditor writing workspace