How to write a research proposal that gets approved

A research proposal is a short document with a disproportionate effect on the next nine months of your life. Approved, it becomes the skeleton of your dissertation and half your methodology chapter is already written. Rejected or heavily revised, you lose weeks at exactly the point when momentum matters most.

The good news is that proposals are assessed against a predictable set of questions, and most rejections come from a short list of avoidable problems. This guide takes each section in turn, explains what the reader is actually checking, and ends with the six reasons proposals come back marked 'revise and resubmit'.

What a proposal is really assessing

A supervisor or panel reading your proposal is answering three questions, in this order. Is this a real question that has not already been answered? Can it be answered with the methods proposed? Can this particular student answer it in the time and with the resources available?

Notice that only the first is about intellectual merit. The other two are about feasibility, and feasibility is where most proposals fail. A brilliant question that requires access to fifty NHS trusts, ethical approval from a research ethics committee and eighteen months of fieldwork is not a good undergraduate proposal, no matter how original it is.

Write with those three questions taped to your monitor. Every section should be visibly answering one of them, and the reader should never have to infer.

The title and the research question

A working title should contain the phenomenon, the population or context, and ideally the approach. 'Hybrid working and perceived autonomy: a qualitative study of mid-career professionals in UK financial services' tells a reader everything in one line. 'An investigation into modern working practices' tells them nothing and signals that the scope has not been decided.

Your research question needs to be answerable, bounded and honest about its type. Descriptive questions ask what is happening; explanatory questions ask why; evaluative questions ask how well something works; exploratory questions ask what is going on in a poorly understood area. Each implies different methods, and stating the type prevents the mismatch that panels notice immediately.

Two or three sub-questions are usually better than one sprawling question, provided they genuinely decompose the main one rather than adding new territory. If a sub-question requires a different sample or a different method, it is a second project.

Justifying the study: literature, gap and contribution

The literature section of a proposal is not a mini literature review. It is a compressed argument in perhaps 600 to 1,200 words that does three things: shows you know the current state of the field, identifies a specific gap, and explains why filling it is worth someone's time.

Be precise about the gap. 'Limited research exists on hybrid working' is not credible in 2026 — there is a great deal. 'Existing UK studies of hybrid working have surveyed employees at a single point in time; none identified here has followed the same individuals across a policy change' is specific, defensible and immediately suggests a design.

Then say what your contribution is, in modest and concrete terms. For undergraduate and master's work, a legitimate contribution can be applying an established framework to a new context, testing a finding in a different population, or bringing qualitative depth to a question previously studied only through surveys. You do not need to promise to change the field.

Methodology: defend choices, don't recite labels

The weakest methodology sections are glossaries. They define positivism and interpretivism, list the differences between qualitative and quantitative research, and then announce a choice without connecting it to the question. The strongest sections barely define anything and spend their words on justification.

Work through the chain in order and give a reason at each link: your philosophical position, your overall design, your sampling strategy and sample size, your data collection instrument, and your analytical approach. Each should follow from the one above and from your question. If your question asks how people make sense of something, a survey with fixed-response items cannot answer it, and a panel will say so.

Be concrete about numbers and logistics, because that is where feasibility is judged. 'Twelve to fifteen semi-structured interviews of roughly 45 minutes, recruited through two professional networks, transcribed and analysed using reflexive thematic analysis following Braun and Clarke' is assessable. 'Interviews will be conducted with a number of participants' is not.

• Design: what kind of study is this, and why does the question require it?
• Sampling: who, how many, recruited how, and what are the inclusion criteria?
• Instruments: what exactly will you ask or measure, and where did the instrument come from?
• Analysis: which named approach, and what will the output look like?
• Quality: how will you address reliability and validity, or credibility and trustworthiness?

Ethics, access and risk

Ethics is the section most often written last and most often the reason for delay. Address the standard four at minimum: informed consent, the right to withdraw and by when, confidentiality and data storage, and any potential for harm or distress including your own as a lone researcher.

Access is the practical twin of ethics and is frequently the real bottleneck. If your design depends on an organisation letting you in, say who, say what stage the conversation has reached, and say what your fallback is if permission does not arrive. A proposal that names a plan B for access is markedly more likely to be approved than one that assumes success.

Where your work involves children, patients, prisoners, personal data or anything commercially sensitive, find out now which approval route applies and how long it takes. In several UK universities the difference between an internal ethics review and an external one is the difference between two weeks and three months, and no amount of good writing recovers that time.

Timeline, resources and word budget

A timeline is a feasibility argument in table form. Break the project into stages with dates, and make sure the total finishes at least three weeks before the submission deadline. Examiners and supervisors both know that everything overruns; a timeline with no slack signals inexperience.

Work backwards from the deadline rather than forwards from today. Submission, final proofread, formatting, complete draft to supervisor, analysis complete, data collection complete, ethics approved, proposal approved. When you do it in that direction the impossible parts of the plan become obvious immediately.

A brief indicative word budget helps too — how many words each chapter is expected to take. It costs you four lines and it shows the reader that you have thought about proportion, which is a surprisingly common weakness in first drafts of dissertations.

Six reasons proposals get sent back

Across supervision meetings, the same handful of problems recur. If your draft has any of these, fix it before you submit rather than after.

• The question is too broad to answer in the word count — usually fixable by narrowing the population or the time frame rather than changing the topic.
• The method cannot answer the question, most often a survey attached to a 'why' question.
• Access has been assumed rather than secured, with no fallback described.
• The gap is asserted, not evidenced — 'little research exists' with three citations behind it.
• The timeline has no slack and no dependency between stages.
• Ethical issues are acknowledged in a single paragraph with no procedure attached.