What goes in a research proposal
Formats differ by funder and department. The underlying questions don't.
The core sections
Each section answers one question the reader has. If a section isn't answering its question, it is the wrong content, however well written.
The proportions below are a starting point. Funders and departments publish their own templates, and those always win.
| Section | Question it answers | Rough share of length |
|---|---|---|
| Title and summary | What is this, in 150 words? | 5% |
| Background and gap | What's known, and what isn't? | 25–30% |
| Research questions or aims | What exactly will you find out? | 5–10% |
| Methods | How will you answer it, and why that way? | 35–40% |
| Timeline, resources, ethics | Can it be done, safely, in time? | 10–15% |
| Expected contribution | Who benefits if it works? | 5–10% |
Turning a topic into a question
Narrow along three lines: who (the population), what (the specific outcome or phenomenon) and where or when (the context). In health research, the PICO format does this formally: Population, Intervention, Comparison, Outcome.
A good question is one where you can picture the table or the quotation that would answer it.
Justify every method choice
For each decision, add a "because". Semi-structured interviews, because the question is about how people reason. A sample of 24, because similar studies reached saturation around that point.
Reviewers forgive modest designs. They rarely forgive unexplained ones.
Be honest about risk
Name the two things most likely to go wrong, such as slow recruitment or a dataset you can't get access to, and say what you would do instead. It reads as competence.