How to write a research proposal, section by section
By Mubarak MutesasiraUpdated 11 min read
Short answer:
To write a research proposal, state one clear research question, show from the literature why it matters and what is still unknown, then explain exactly how you will answer it: design, sample, data, analysis, ethics, timeline and budget. Write the methods before the introduction, keep every section tied to the question, and finish inside the word or page limit you were given.
This guide is for students writing a dissertation or PhD proposal and early-career researchers writing their first funded proposal. It is the hub of our series on proposals: you can follow it with a full research proposal example, APA formatting for a proposal and how to write an IRB application.
What is a research proposal?
A research proposal is a document that describes a study you plan to carry out, argues that the question is worth answering, and shows that your method can answer it with the time and resources you have. Readers use it to make a decision: admit you to a program, approve your dissertation topic, or fund the work.
Every reader of a proposal is asking three questions:
- Is the question worth answering? (significance)
- Can this method answer it? (rigor)
- Can this person do it, here, in this time? (feasibility)
If a paragraph helps none of these, cut it.
What are the parts of a research proposal?
Most proposals use the same ten parts, though templates rename and merge them. The table gives the purpose of each part and a typical share of a 3,000-word proposal.
| Section | What it answers | Typical share | Words in a 3,000-word proposal |
|---|---|---|---|
| Title | What is the study, in one line? | n/a | 10 to 20 |
| Abstract or summary | The whole proposal in brief | 5% | 150 to 250 |
| Introduction and problem | Why does this matter now? | 10% | 300 |
| Research question, aims, objectives | What exactly will you find out? | 5% | 150 |
| Literature review | What is known, and what is the gap? | 25% | 750 |
| Methods | How will you answer the question? | 30 to 35% | 950 |
| Ethical considerations | Who could be harmed, and how will you prevent it? | 5% | 150 |
| Timeline | When will each step happen? | 5% | 150 (often a table) |
| Budget | What will it cost? (funded work only) | 0 to 5% | 0 to 150 |
| References | Which sources support your claims? | Not counted in most limits | n/a |
The methods section is the largest because it is where most proposals fail. Reviewers forgive a modest literature review. They rarely forgive a method that cannot answer the question.
How long is a research proposal?
There is no single length. The limit comes from whoever reads it.
| Context | Typical length | Who sets it |
|---|---|---|
| Undergraduate dissertation proposal | 1,000 to 2,000 words | Course handbook |
| Master's dissertation proposal | 1,500 to 3,000 words | Program handbook |
| PhD application (UK, Europe, Australia) | 1,500 to 3,000 words, often 2 to 4 pages | Department admissions page |
| US PhD candidacy or prospectus | 10 to 30 pages, sometimes the first three chapters | Graduate school and committee |
| NIH R01 grant | 1-page Specific Aims plus a Research Strategy commonly limited to 12 pages [1] | NIH and the funding announcement |
| NSF grant | 1-page Project Summary plus a Project Description of up to 15 pages [2] | NSF PAPPG and the solicitation |
Student ranges are typical, not rules: your handbook or the funding call overrides them.
How to write a research proposal in 8 steps
Write the sections out of order: settle the method early, because it constrains everything else.
- Write the question in one sentence. Use FINER (feasible, interesting, novel, ethical, relevant) to test it. A PICO form helps for clinical and intervention questions. See how to write a research question.
- Read until you can name the gap. Collect 20 to 40 key papers. Stop when new papers repeat what you know and you can say in one sentence what nobody has tested yet.
- Draft the methods. Design, participants or data, sampling and sample size, instruments, procedure, analysis. Check that each objective has a matching analysis.
- Write the aims and objectives. One aim, then two to four objectives that start with a verb you can check: measure, compare, describe, test.
- Write the literature review as an argument that ends at your gap, not a list of summaries.
- Write the introduction and significance last among the main sections, because now you know what you are introducing.
- Add ethics, timeline and budget. Build the timeline backwards from the submission date and include ethics approval time.
- Write the title and abstract, then cut. Read the whole thing against the limit and the marking criteria or review criteria.
What goes in each section?
Title
A good title names the topic, population and design, in under 20 words: "Weekly retrieval quizzes and exam performance in introductory statistics: a quasi-experimental study."
Abstract or summary
A proposal abstract is a 150 to 250 word summary of the problem, question, method and expected contribution. It is written in the future tense for what you will do. Many student proposals do not need one; check your template.
Introduction and problem statement
Open with the problem, not the history of the field: what it is, who it affects, what has been tried and why the gap matters. End with the research question.
A problem statement is a short passage that names a specific, evidenced problem and the consequence of leaving it unsolved. "Pass rates in introductory statistics at our institution were 64% over the last three years" is a problem statement. "Statistics is hard for many students" is not.
Research question, aims and objectives
The research aim is the overall purpose of the study; objectives are the specific, checkable steps that achieve it. If you test a prediction, add hypotheses here in null and alternative form. Each objective should map to one analysis in the methods.
Literature review
Organize by theme or debate, not by author. Each paragraph should make a claim about the literature ("Classroom studies replicate the laboratory effect, but almost all use multiple-choice quizzes") and cite the studies behind it. Close with the gap and how your study fills it.
Methods
Cover, in this order: design and why it suits the question; setting and participants; sampling and sample size (with a power calculation for quantitative work or a justification for qualitative work); instruments and measures; procedure; analysis plan, naming the tests; validity, reliability or trustworthiness; limitations. Write in the future tense.
Ethical considerations
Name the risks to participants, how you will get informed consent, how you will store and protect data, and which ethics board will review the study. For human participants, approval usually has to arrive before data collection starts. The IRB application guide covers this in detail.
Timeline
A table or Gantt chart with phases and milestones by month. Include ethics review, recruitment, analysis and writing, and leave a buffer of about 10 to 15% of the time.
Budget
Only for funded work or where asked. List each cost with a one-line justification: participant payments, equipment, software, travel.
References
Cite every claim and every borrowed method in the style your program requires, often APA 7. See research proposal APA format.
How does a PhD proposal differ from a master's proposal?
A PhD proposal must show an original contribution across three or more years; a master's proposal must show a well-designed study that fits in months. The structure is the same, but the weight shifts.
| Master's proposal | PhD proposal | |
|---|---|---|
| Contribution | Applies or tests known methods well | Adds new knowledge to the field |
| Scope | One study, 6 to 12 months | Two to four linked studies or a program of work |
| Literature | Enough to justify the question | Shows command of the field and its debates |
| Feasibility | Can you finish by the deadline? | Can you finish, and do you fit this supervisor's expertise? |
| Common format | 1,500 to 3,000 words | 2,000 to 4,000 words for applications; longer for candidacy |
How do grant proposals differ by funder?
Grant proposals follow the funder's headings, not yours. The questions are the same, but each funder asks them in its own order and limits.
- NIH leads with a one-page Specific Aims page, then a Research Strategy organized as Significance, Innovation and Approach. [1]
- NSF requires a one-page Project Summary with Overview, Intellectual Merit and Broader Impacts, and a Project Description with a separate Broader Impacts section. [2]
- ERC grants are split into an Extended Synopsis and CV for the first step and a full Scientific Proposal for the second. [3]
- Wellcome schemes use online forms with their own question sets; read the scheme page before drafting. [4]
Always paste the call's assessment criteria next to your draft and check that each criterion has a paragraph that answers it.
What are the most common faults in research proposals?
Most rejected proposals fail on scope or method, not on writing. These are the faults reviewers and supervisors mention most often, with a fix for each.
| Fault | What it looks like | Fix |
|---|---|---|
| Question too broad | "How does social media affect mental health?" | Name a population, exposure, outcome and setting |
| No gap | A literature review that summarizes papers and stops | End the review with one sentence on what is not known |
| Method does not match question | A causal question answered with a one-off survey | Choose a design that can support the claim you want to make |
| No sample size reasoning | "About 100 participants" | Give a power calculation or a qualitative sufficiency argument |
| Unrealistic timeline | Recruitment, analysis and writing in three months | Build backwards from the deadline and add ethics review time |
| Ethics as an afterthought | One line: "Ethics approval will be obtained" | Name the risks, consent process, data storage and the board |
| Overrun | 4,200 words against a 3,000-word limit | Cut background first; never cut methods to fit |
How to write a research proposal in AI for Research
AI for Research drafts proposals from your topic, with background sections that cite real papers. You still choose the question and methods, and check every citation.
- Proposal Studio, Research proposal: the whole proposal (15 credits) or one section (3 credits): aims, background, literature review, methods, timeline, budget outline or ethics. Budget amounts come back as placeholders.
- Grant proposals in generic, NIH, NSF, ERC or Wellcome format, with the funding call pasted in, at the same costs. Try the grant proposal writer.
- Thesis Suite for the groundwork: Research question (FINER, PICO, PICOT, SPIDER), Hypotheses, Methodology builder, Research gaps and Timeline, 3 credits each.
- Find Papers and the Library for background reading (0 credits). Share the proposal's Project with your supervisor as "can comment".
- Reviewer set to "proposal" scores each criterion out of 10 and quotes passages with a fix. On the free plan, a preview of the first 3,000 words (the full scorecard and top 3 comments) costs 3 credits.
The free research proposal generator runs 3 times a day without an account. The free plan's 45 credits a month cover one whole draft and several section rewrites; see pricing. No tool can judge whether your idea is new or feasible at your institution. Your supervisor can.
FAQ
What are the parts of a research proposal?
Most research proposals have a title, a short abstract or summary, an introduction with the problem and research question, aims and objectives, a literature review, methods, ethical considerations, a timeline, a budget where funding is involved, and references. Universities and funders rename and merge these, so always check the template you were given, but the questions each part answers stay the same.
How long is a research proposal?
It depends on the purpose. A PhD application proposal is often 1,500 to 3,000 words. A master's or undergraduate dissertation proposal is often 1,000 to 3,000 words. Grant proposals are set in pages by the funder: an NIH R01 Research Strategy, for example, is commonly limited to 12 pages. The limit in your program handbook or funding call always wins.
What is the difference between a research proposal and a thesis?
A proposal describes research you plan to do and argues that it is worth doing and feasible. It is written mostly in the future tense and has no results. A thesis reports research you have done, with results, discussion and conclusions. A good proposal often becomes the first draft of the thesis introduction, literature review and methods chapters.
How long does it take to write a research proposal?
For a PhD application, plan four to eight weeks of part-time work: two to three weeks of reading, one week to settle the question and methods, and two to three weeks of drafting and revision with feedback from a potential supervisor. A grant proposal for a large funder often takes two to three months, because budgets and institutional sign-off add time.
Can AI write a research proposal for me?
AI can draft sections, suggest structure and find papers for the background, but you must supply the idea, choose the methods and check every citation. Committees judge whether you understand your own project, and a proposal you cannot defend in an interview will fail. Use AI for a first draft or a review pass, then rewrite it in your own words.
Next steps
- Read a research proposal example with notes on every section.
- Format it with research proposal APA format.
- Prepare your ethics submission with how to write an IRB application.
- Sharpen the question first with how to write a research question, and plan the method with how to write a methodology chapter.
Sources
- National Institutes of Health. Page limits (Table of Page Limits). https://grants.nih.gov/grants-process/write-application/how-to-apply-application-guide/format-and-write/page-limits
- National Science Foundation. Proposal & Award Policies & Procedures Guide (PAPPG), Chapter II: Proposal Preparation Instructions. https://www.nsf.gov/policies/pappg
- European Research Council. Work Programme and Information for Applicants. https://erc.europa.eu/apply-grant
- Wellcome. Funding schemes. https://wellcome.org/research-funding/schemes
- AI for Research. Research proposal generator. /tools/research-proposal-generator