PhD Research Proposal Example: UK, 1,500 Words

PhD Research Proposal Example: UK, 1,500 Words

Plan a UK PhD research proposal to 1,500 words unless your department says otherwise. Manchester calls 1,000 to 1,500 typical, one UWE studentship sets 1,500 with the bibliography outside the count, and departments at York alone range from 500 to 4,000.

At 1,500 words, the literature review is the wrong place to spend most of your space. A short UK proposal is read for whether the project is clear, original and doable in three years, so most of the words belong to questions, method, access and timeline. Below is a section-by-section budget and a full constructed proposal written to it, annotated against what UK universities say they look for. If you already have a draft, a PhD research proposal review checks those same sections.

Every university page quoted below was checked on 29 October 2026. Most give no year or cycle, and departments change their rules, so recheck your own department's page before you submit. The PhD research proposal guide for Oxford, Cambridge, LSE and ETH covers those universities' own routes; this page is the worked example.

Contents

• How long a UK PhD proposal is
• What the reader is checking
• The 1,500-word budget
• Constructed example: a full proposal, annotated
• What to cut when you are at 2,100 words
• Before you send it to a supervisor
• Questions about a 1,500-word PhD proposal

How long a UK PhD proposal is

There is no national rule. Each university, and often each department, sets its own. Here are 13 published rules from seven UK universities, quoted as written:

Our arithmetic: 1,500 words falls inside, or at the edge of, 8 of these 13 rules. The other five sit below it (York History, York Health Sciences) or well above it (York Social Policy, York Management, Surrey). That is why 1,500 is the number to plan for, and why it is not a number to assume.

The column that changes your draft most is the last one. At one university, York, the English department counts references inside 1,500 words, Law excludes the bibliography, and History excludes footnotes with sources. A reference list of 20 items can run to 300 or 400 words (our estimate, at 15 to 20 words per entry). Under York English's rule, those words come out of your argument.

Our view: settle the reference rule before you write a single section. If your department does not say, ask it. If you cannot get an answer, write the proposal so that it fits the limit with the references included.

What the reader is checking

Manchester's guidance says a proposal should:

"demonstrate evidence of intellectual purpose and originality; show that you are capable of communicating your ideas clearly, concisely and coherently; define the topic you are interested in and show good awareness of the research context."

Its list of sections asks direct questions about method and plan: "What form will the resources take? Where are they located? Will there be any problems of access?" and "What are the main project stages? What would the annual completion expectations be? What are the perceived challenges, and how will these be overcome?"

UWE's studentship guidance asks you to "Justify that what you propose to undertake can be completed within the duration of your registration period" and to "Confirm you have provisional permission to undertake the research on the material, if permission to do so is required." On the bibliography it is blunt: "It is not meant as an opportunity for you to list all the works in your field with which you are familiar."

Put together, the published guidance gives you a seven-point checklist:

1. A working title that says what the project is.
2. A problem or question a reader could quote back to you.
3. Enough context to show where the question sits, and what is not yet known.
4. A method that could actually answer each question.
5. Named materials or data, and how you will get access.
6. A timeline that ends inside the doctorate.
7. A stated contribution: what will exist at the end that did not before.

Only item 3 is literature. The other six are about the project.

The 1,500-word budget

Our recommendation, not a university rule. None of the pages above gives a split.

Our arithmetic: 150 + 350 + 150 + 450 + 50 + 200 + 150 = 1,500. Literature gets about a quarter. Method, access and timeline together get 650 words, or 43%.

The reasoning: a 1,500-word cap squeezes method and feasibility first, because applicants write the literature review first and run out of room. UWE asks for a structure indication ("a brief indication of what this structure currently looks like"), which is why the budget keeps 50 words for it.

The countercase: in history, literature and other text-based humanities, the "method" is often a reading of sources and a theoretical approach, and the argument with existing scholarship is itself the contribution. There the context section can reasonably take 450 to 500 words, with method shrinking to match. UWE's own list allows for this: "the methodologies for your work and any relevant theoretical approaches to be deployed".

Constructed example: a full proposal, annotated

Constructed example, written by us, not a real applicant's proposal. The city of Halverton is invented, and so are the pilot results and the council's agreement. Citations are placeholders in square brackets: the literature claims are illustrative, not a review of the real field, and a real proposal must source each one. The bibliography sits outside the count, as in UWE's guidance.

The proposal is a public-policy PhD in the social sciences. It is 1,468 words by our count, excluding the section labels in the headings below and the bibliography.

Title and research problem (151 words)

Working title: Heat on the agenda: how one English city council turned heat warnings into decisions, 2016–2026

Halverton City Council in England receives heat alerts and climate risk assessments, and it is expected to act on each one. Yet a warning is not a decision. Between the moment a heat alert or a climate risk assessment reaches a council and the moment something changes on the ground, such as a cooling centre opening or a planning condition on shade, there is a chain of meetings, reports and choices that is rarely studied directly. This project traces that chain in one city, Halverton, over ten years. Halverton is a useful case because it declared a climate emergency in 2019 and publishes its committee papers for the whole period. The project asks which warnings became decisions, which did not, and why, and it treats the council's own paper trail as the main evidence.

Annotation. Budget 150, actual 151. The title names the place, the period and the process. The last sentence is the question a reader can quote back (checklist 2). The reason for choosing the case is one sentence, not a paragraph.

Context and literature (356 words)

Three bodies of work frame the project. The first, on climate adaptation in local government, has mapped what councils say they will do in adaptation plans and strategies [Author A, 20XX; Author B, 20XX]. It is less clear how those plans connect to the routine committee decisions that spend money and set rules. The second, on public-health emergency planning, treats heat mainly as a short-term response problem: alerts, vulnerable groups and services during a hot spell [Author C, 20XX]. That work explains what happens in the week of a heatwave but says little about longer-term decisions on buildings, trees, housing standards or planning, which are made months later by different people. The third, on policy uptake, studies how evidence enters decisions and names conditions under which it is used, delayed, reframed or ignored, such as the timing of the evidence, its cost implications, the standing of the person who brings it and its fit with existing priorities [Author D, 20XX; Author E, 20XX].

Each literature looks at a different stage of the same process, and none follows a single risk from warning to decision inside one organisation over a long period. That is the gap this project addresses, and it is a bounded one. I do not claim that no one has studied heat in local government. I claim that the step from a received warning to a recorded decision has not been traced across a whole council over a decade, using the council's own records alongside the people who made the decisions. A single case cannot show how common each pattern is across England, but it can show the mechanism in enough detail for later comparative work to test.

This matters for two reasons. Practically, if most warnings stop at one stage, such as an officer report that never reaches a committee, then better forecasts alone will not change outcomes, and councils need to fix the stage where warnings stall. Conceptually, the project tests whether the uptake conditions identified for national policy also hold in a local authority, where budgets are smaller, officers carry several briefs at once and decisions are spread across many committees.

Annotation. Budget 350, actual 356, about 24% of the whole. Three literatures, each given one job, then the gap. The gap is bounded: the proposal says what it does not claim. This answers Manchester's "show good awareness of the research context" without becoming a literature review.

Aims and research questions (146 words)

The aim is to explain how heat warnings become, or fail to become, council decisions in one English city. Three questions guide the research:

RQ1. Which heat-related warnings reached Halverton City Council between 2016 and 2026, through which channels, and at which point in the council's process did each one stop or turn into a decision?

RQ2. What distinguished the warnings that became decisions from those that did not, in terms of their source, timing, cost and the committee that received them?

RQ3. How do officers and councillors explain those outcomes, and do their explanations match the record?

The project does not judge whether the council's decisions were correct. Its scope is the decisions the council recorded, not informal action that left no trace in its papers. Together the three questions move from what happened, to what predicted it, to how the people involved understood it.

Annotation. Budget 150, actual 146. Each question has a method in the next section, and the closing scope sentence stops the project from growing.

Methods, materials, access and ethics (438 words)

The design is a single-city case study using two kinds of evidence: the council's published records and interviews.

Documents (RQ1, RQ2). Halverton publishes committee agendas, officer reports and minutes online for the whole period. I will build a register of every document from 2016 to 2026 that mentions heat, high temperatures or overheating, then code each item for its source (national alert, consultant report, public-health team, resident petition), the committee or officer that received it, and what happened next: noted, deferred, referred, funded or adopted. A pilot search of two years of minutes, which I ran in preparing this proposal, returned 41 relevant items, so I expect a register of a few hundred items over ten years. That is large enough to compare and small enough for one researcher to code by hand, using qualitative analysis software the university already licenses. To test the coding frame, a second coder will code a 10% sample, and I will revise any category where we disagree often.

Interviews (RQ3). I will conduct around 25 semi-structured interviews with current and former officers and councillors who handled heat-related items, chosen from the register so that both adopted and stalled items are covered. Interviews will ask participants to explain specific items from the register, which keeps them tied to the record rather than to general opinions. Each interview will last about an hour and will be recorded and transcribed with consent.

Analysis. RQ1 uses the coded register to map where items stopped. RQ2 compares adopted and stalled items across the coded variables, first descriptively and then with a simple comparison of proportions across sources and committees. RQ3 compares interview explanations with the documented sequence for the same items, which is a direct check on whether participants' accounts fit the record.

Training. I used document coding in my master's dissertation, but I have not run research interviews. I will take the university's qualitative interviewing course in the first year and pilot the interview guide with two former officers before the main round.

Access and ethics. The documents are public, so the register raises no access problem and needs no permission beyond normal research ethics. Interview data will be stored on university servers only. For interviews, the council's democratic services team has agreed in principle to circulate an invitation to current staff; former staff and councillors will be contacted through public channels. If fewer than 15 people agree to take part, I will narrow RQ3 to the two committees with the most items rather than drop it. All interviews require university ethics approval, informed consent and anonymisation, and no participant will be named in the thesis.

Annotation. Budget 450, actual 438, the largest section. It answers Manchester's three access questions (what form, where, any problems), names a fallback for the main risk, and admits a skills gap with a training plan. Each research question is mapped to a method. For more on these two moves, see proposing PhD methods you have not used yet and proving data access is feasible.

Provisional structure (50 words)

Provisional structure, in six chapters: introduction and literature; methods; the register of warnings (RQ1); what separates adopted from stalled items (RQ2); how decision-makers explain the record (RQ3); and conclusions for councils and for uptake research. The chapter order follows the research questions, so each findings chapter answers one of them.

Annotation. Budget 50, actual 50. UWE asks for "a brief indication of what this structure currently looks like". One sentence per chapter would be too long at this cap; a list is enough.

Timeline and risks (190 words)

The project is designed to fit a three-year full-time doctorate.

Year 1: complete the literature review, finalise and test the coding frame, take the interviewing course, obtain ethics approval and build the document register for 2016 to 2026. The register is the foundation for everything else, so it is scheduled first and does not depend on interview access.

Year 2: code the register, run the RQ1 and RQ2 analysis, and conduct the interviews in two rounds, so that the second round can follow up on patterns found in the first. Present the register findings at one conference.

Year 3: analyse the interviews against the record (RQ3), write the remaining chapters, share a draft briefing with the council, and submit by the end of the third year.

The main risk is interview recruitment. It is managed in two ways: the document work in years 1 and 2 can be completed without any interviews, and the fallback for low recruitment is already defined above. A second risk is that the council changes how it publishes records, so I will download and archive the full set of documents in the first three months.

Annotation. Budget 200, actual 190. Year by year, with the work that does not depend on other people scheduled first. It answers Manchester's "What are the perceived challenges, and how will these be overcome?"

Contribution and outcomes (137 words)

The project will produce three things. The first is a decade-long register of how one council handled heat warnings, which other researchers can reuse. The second is a tested explanation of why some warnings become decisions and others do not. The third is a comparison of official records with participants' own accounts of the same decisions. For research, it extends uptake theory from national policy to local government, and it offers a method, the warning register, that can be repeated in other councils. For practice, it gives councils a simple way to audit their own handling of warnings: list them, trace them, and find the stage where they stop. Beyond the thesis, I plan a short briefing for Halverton and for local-government networks, and at least one journal article from the register analysis, submitted before the thesis.

Annotation. Budget 150, actual 137. Three outputs, named. UWE asks for "any outcomes you envisage beyond the PhD thesis itself", which the last sentence answers.

Our arithmetic: 151 + 356 + 146 + 438 + 50 + 190 + 137 = 1,468, which leaves 32 words under a 1,500 cap. Questions, methods and timeline (146 + 438 + 190) take 774 words, or 53%.

What to cut when you are at 2,100 words

If your draft runs long, cut in this order, and stop as soon as you are under:

1. Move the bibliography out, if the rules allow. Check the reference rule first (see the table). Where references are excluded, this is free.
2. Cut literature summary before anything else in the text. Keep one sentence per body of work and one sentence on the gap. Delete descriptions of individual studies.
3. Move supervisor and fit material to its own field. UWE's studentship asks for "Choice of supervisors – 300 words" separately. When your application has a field like that, the proposal should not repeat it.
4. Merge the structure into the timeline if you are still over: one line per year that names the chapter it produces.
5. Do not cut the timeline or the access plan. They are what UWE asks you to "Justify" and what Manchester asks about directly.

If you are still over after step 4, the project is probably too big for the cap, not too wordy. Narrow a research question.

Before you send it to a supervisor

Manchester tells applicants to "Contact an academic member of our staff to discuss your research proposal and key objectives before you submit your formal application," to "fine-tune your proposal and check that we can identify a suitable supervisory team for you." It also says that "you will not be committing yourself to the precise detail or methodology."

That last sentence is a reason to be specific, not vague. A precise plan that later changes is normal. A vague plan gives a supervisor nothing to respond to. When feedback comes back, the guide to using PhD supervisor feedback well covers how to revise without losing your own project.

Questions about a 1,500-word PhD proposal

Does the bibliography count toward 1,500 words?

It depends on the department, and the answer can differ inside one university. At York, English counts references within 1,500 words, Law excludes the bibliography, and History excludes footnotes with sources. UWE's DCRC guidance puts the bibliography outside the count. If your page is silent, ask, or fit the limit with references included.

Should I include a Gantt chart?

Not usually at 1,500 words. Exeter's guidance says: "You don't need to produce a detailed time plan, but it is helpful to provide a summary of what you are planning to do and when." A year-by-year paragraph, like the one in the example, does that job in under 200 words. Use a chart only if your department asks for one.

Do I need a hypothesis?

Not necessarily. Exeter says research objectives "could be articulated as hypotheses, propositions, research questions, or problems to solve". Use a hypothesis if your field and method test one. The example above uses research questions because its method is a case study.

Can I send the same proposal to two universities?

The project can travel; the fit cannot. Lengths and reference rules differ (see the table), and some departments ask why you chose them: York English asks you to state "the academic factors, eg facilities, resources or staff, which have led you to apply to York". Keep the core question and method, and rewrite the length, the fit and anything a supervisor there asked you to change.

Is a research proposal the same as a statement of purpose?

No. The proposal is about the project. A statement of purpose is about you. Research statement vs SOP vs research proposal sets out what each document does. US programs that ask for a report on past work rather than a proposal want a different document; see this research statement example.

Applying in Australia? The first document there is shorter: see our Australian PhD expression of interest example, with an outline on Adelaide University's section limits.

Get a free review of your PhD research proposal

Get a free Quick Review of your PhD research proposal against GradPilot's PhD research proposal rubric for UK, Irish and Australian applications. It reads for a clear title and question, literature that defines what is left open, methods that answer each question, a timeline that fits the doctorate, resources, access and ethics, and the contribution. If your application combines a proposal with a statement of purpose, use the PhD research proposal and statement of purpose rubric; for a separate supervisor-choice answer, use the PhD adviser-fit rubric. The PhD statement of purpose review covers the SOP itself.

More PhD proposal guides:

• PhD research proposal guide: Oxford, Cambridge, LSE and ETH: what those universities require.
• UQ PhD expression of interest and proposal: how the proposal fits an Australian supervisor-first route.
• PhD essays: every PhD application guide on GradPilot.

Sources

Retrieved 29 October 2026, except where noted.

• University of Manchester, How to write a research proposal (no year stated): typical length, assessment points, section questions, supervisor contact, refinement after admission.
• UWE Bristol, DCRC PhD research project proposal guidance (2022 studentship round, one research centre): 1,500-word proposal, contents, bibliography outside the count, 300-word supervisor field.
• University of York, Research proposals (no year stated): department-by-department lengths and reference rules.
• University of Aberdeen, School of Language, Literature, Music and Visual Culture, How to write a research proposal (no year stated).
• University of Sheffield, Sociological Studies, Politics and International Relations, Writing a research proposal (no year stated).
• University of Exeter, Research proposal (no year stated): length guide, time plan, hypotheses.
• University of Surrey, PhD proposal guidance (no year stated): 3,500 to 5,000 words for named subjects.