A research proposal is a promise and a plan. Learn the sections reviewers expect, how to frame a sharp question, write a methodology that survives scrutiny, and the feedback loop that turns a draft into an approval.
Every thesis, funded project and PhD place begins with the same small document: the research proposal. It is usually only a few pages long, yet it carries an unreasonable amount of weight — it is the thing a supervisor reads before agreeing to supervise you, the thing a committee reads before approving your topic, and the thing a funding body reads before trusting you with money. Most students write it backwards: they describe their topic at length and hope the reader finds the point. Reviewers, who read dozens of these, do the opposite — they hunt for the point first, and if they cannot find it within a few minutes, the proposal is in trouble. This guide shows you how to put the point where they look for it.
Final-year undergraduates writing a dissertation proposal, master's students applying for thesis approval, and anyone preparing a proposal for a research grant or a PhD application. The examples lean academic, but the underlying structure — question, evidence of a gap, method, plan — is what any reviewer in any field is looking for.
A proposal is not a mini-thesis. Nobody expects results; they expect a convincing promise. Think of it as two things stapled together: a pitch (why this question deserves an answer) and a plan (how you will answer it, with what, and by when). Every reviewer, consciously or not, is asking three questions as they read:
Your proposal is approved when all three answers are yes. Every section exists to push one of those answers toward yes.
Formats differ — your department or the funding call will give you headings and page limits, and those win every argument. But underneath the headings, the logic is always the same. Here is the standard skeleton and the job each part does:
| Section | The reviewer's question | Common mistake |
|---|---|---|
| Working title | What is this about? | A clever title that hides the topic — be descriptive first |
| Abstract / summary | Should I keep reading? | Written first, so it promises things the proposal never delivers |
| Introduction & research questions | What exactly will you find out? | A broad topic with no sharp, answerable question |
| Literature review / background | What is missing that you will add? | A book report that summarises everything and concludes nothing |
| Methodology | How will you do it, exactly? | Vague verbs — 'analyse', 'explore' — with no data, tools or steps |
| Timeline | Is this feasible in the time given? | Twelve months of work squeezed into six |
| Expected outcomes & significance | Who benefits, and how? | Grand claims the method cannot support |
| References | Have you read the field? | A thin or outdated list that signals shallow preparation |
Not every call asks for every section — but if a section is missing, its question still gets asked. Answer it somewhere.
Most weak proposals start with a topic: 'social media and mental health', 'renewable energy in rural areas'. Topics are not questions, and reviewers cannot approve a topic — there is nothing to say yes to. A question names the unknown: 'Does a two-week break from image-based social apps reduce anxiety scores in undergraduates?' Now there is something to investigate, a method to design, and a result to report.
The test of a sharp question is whether you can state, in one sentence, what would count as an answer. If you cannot, the question is still a topic wearing a question mark. Sharpen it by adding the who, the what and the where until the sentence snaps into focus. 'How do teachers feel about AI?' becomes 'How do secondary-school science teachers in my city use AI tools when planning lessons, and what stops them using those tools more?' The second version tells the reviewer exactly what data you will collect and from whom — which is half the methodology already written.
The literature review is the most misunderstood section. It is not a summary of everything ever written on your topic — reviewers already know the field. Its job is cartography: sketch the territory, mark what is known, and point at the blank patch your research will fill. Every paper you cite should earn its place by doing one of three things: establishing a fact you build on, showing a method you borrow or improve, or demonstrating the gap you will close.
The section should end where your research begins. A single clear sentence — 'No study has yet examined X in a Y context using a Z method' — hands the reviewer the reason your proposal exists. If you cannot write that sentence, you have not found your gap yet, and no amount of elegant prose will cover it.
Write your gap sentence before you write the literature review. Tape it above your desk. Every paragraph of the review should be marching toward it — anything marching elsewhere gets cut.
Reviewers forgive an ambitious question; they rarely forgive a vague method. This is the section where proposals live or die, because it answers the hardest of the three questions: can this person actually do it? Vague verbs are the tell — 'data will be analysed', 'participants will be explored'. Replace every vague verb with a concrete noun: which data, how many, collected how, analysed with what.
Name your design plainly (survey, interviews, experiment, case study, archival analysis), then walk the reader through the chain: sources or participants, collection procedure, analysis steps, and the limits you already see. Limits do not weaken a proposal — pretending there are none does. A short, honest paragraph on ethics (consent, privacy, approvals you will seek) signals a researcher the committee can trust.
A timeline is a feasibility argument in disguise. Break the work into phases — preparation, data collection, analysis, writing — and assign each a slice of the available time. Reviewers have seen hundreds of timelines; they can smell a fantasy. Put the buffer where things actually slip (recruitment, approvals, transcription), and keep the writing phase generous — every researcher underestimates writing.
No proposal should meet its reviewer before it has met a friendly reader. Your supervisor sees the field; a friend outside your field sees whether a stranger can follow it. Give them the draft with one instruction: 'Tell me where you got lost or bored.' Those are the exact places the reviewer will stall. Revise, then read the whole thing aloud once — clumsy sentences hide in silence and reveal themselves in speech.
There is no universal page count — a dissertation proposal might be five pages, a grant application thirty. The only rule is to follow the call: its headings, its word limits, its reference style. Reviewers notice compliance because non-compliance wastes their time. When no format is given, the table above is a safe skeleton, and clarity beats length every time.
A proposal is a small document with one job: to make a busy, sceptical reader believe in your question and your ability to answer it. Start with the sharp question, show the gap, describe the method as if someone must repeat it, and let a friendly reader find the weak spots before the reviewer does. Do that, and approval stops feeling like luck.
As long as the call or your department asks for — that is the only number that matters. Dissertation proposals are often 1,500 to 3,000 words; grant proposals can run much longer. When no length is given, aim for the shortest document that still answers the reviewer's three questions: is it worth doing, can you do it, and what will it produce.
Usually yes, within reason — proposals are plans, not contracts, and research questions often sharpen once work begins. Small refinements are normal and expected. A wholesale change of topic normally needs your supervisor's or committee's agreement, so keep them in the loop early rather than presenting a surprise.
Enough to prove the gap. That usually means the key studies that define the current state of knowledge on your specific question — often a dozen or two well-chosen sources rather than a hundred skimmed ones. If every citation is marching toward your gap sentence, you have enough; if you are citing to look thorough, you have too many.
Take it seriously — they know the field's dead ends. Ask what specifically worries them: feasibility, originality, or scope. Often the question survives in a narrower version both of you can support. A supervisor who argued with your proposal is a supervisor who will defend it later.
For most student proposals, no — but a small pilot strengthens any proposal enormously. Even five trial interviews or a quick analysis of a sample dataset proves your method works and your access is real. If you have pilot data, give it a short section; reviewers read it as evidence you have already started.
Keep reading
Your thesis topic will live with you for months, so choose it properly. Learn how to brainstorm widely, test ideas against interest, feasibility and contribution, and narrow a vague interest into a question you can actually finish.
Research papers are not novels. Learn the three-pass method — skim, read, interrogate — with time budgets, a comparison table and the five-question note that makes papers stick.
Set up Zotero from scratch: install the connector, save references in one click, organise collections, and generate a perfectly formatted bibliography.