How to Write a Research Proposal That Starts With the Right Question

Many students sit down to write a research proposal and start with the formatting: headings, page counts, reference style. That is the wrong end of the task. A proposal is not a form to fill in. It is a short, persuasive argument that says: here is a question worth answering, and here is how I plan to answer it. Everything else follows from the question. This guide walks you through building that question and the proposal around it, one calm step at a time.

Start with a question, not a topic

A topic is a broad area, such as “remote work” or “bilingual education.” A question is something you can actually investigate and answer. Reviewers want the question, because a question shows you have a focus.

Compare these:

• Topic: Remote work and productivity.
• Question: Does working from home two days a week affect the on-time completion of weekly tasks for office staff in mid-sized companies?

The second version is narrower, and that is its strength. It names who you are studying, what you are measuring, and the rough boundaries of the study. A good question is specific, answerable within your time and resources, and genuinely open (you do not already know the answer).

A quick test: if you can answer your question with a single web search, it is too small. If you could spend ten years on it, it is too big. Aim for the middle.

Justify why the question matters

Once you have a question, the next job is to explain why anyone should care. This is the part reviewers use to decide whether your project is worth supporting. You are answering three quiet questions:

1. What do we already know? Briefly summarise the existing thinking on the topic.
2. What is missing or unsettled? Point to the gap, contradiction, or unanswered piece.
3. What will your study add? Connect your question directly to that gap.

Keep this short and honest. You are not reviewing everything ever written; you are showing that your question sits in a real gap. Avoid grand claims like “no one has ever studied this.” Instead, write something measured: “Existing studies focus on large corporations, so it remains unclear how these patterns apply to smaller firms.”

Choose a method you can actually carry out

The method section explains how you will answer your question. The most common mistake here is ambition that outruns reality. Choose an approach that fits your time, skills, and access.

Decide and state plainly:

• What kind of evidence answers your question — numbers, interviews, documents, observations?
• Where it comes from — a survey, a set of texts, public records, a small group of participants?
• How you will gather it — a questionnaire, structured interviews, a coding scheme for documents?
• How you will make sense of it — basic statistics, thematic analysis, comparison?

Tie each choice back to the question. If your question asks whether something happens (how often, how much), you likely need numbers. If it asks how or why people experience something, you likely need words. Matching the method to the question is what makes a proposal feel coherent.

A worked example

Here is a short, realistic skeleton you can adapt. Notice how every section answers to the question at the top.

Working title: Two-day remote work and weekly task completion in mid-sized firms

Research question:
  Does a two-day-per-week remote schedule affect the on-time
  completion of weekly tasks among office staff in mid-sized companies?

Why it matters:
  Most remote-work research studies large corporations; the effect on
  smaller firms with fewer support systems is less clear.

Method:
  - Approach: short survey + completion records from one firm
  - Participants: ~40 office staff, one mid-sized company
  - Data: weekly task-completion logs (8 weeks) + a brief questionnaire
  - Analysis: compare on-time rates on remote vs. in-office weeks

Timeline:
  Weeks 1-2  Finalise survey and gain permission
  Weeks 3-10 Collect completion data
  Weeks 11-12 Analyse and write up

This fits on a page, yet a reviewer can immediately see what you will do and why. That clarity is the whole point.

Build a realistic timeline and scope

Reviewers are reassured by a plan that admits limits. A proposal that promises to interview 500 people in a month signals trouble. Break your project into phases with rough dates, as in the example above, and leave room for the parts that always run late: getting permissions, recruiting participants, and writing up.

It also helps to name your scope honestly. State what your study will not cover. Writing “this study examines one firm and does not aim to generalise to all companies” is a strength, not a weakness. It shows you understand the boundaries of your own work.

Common mistakes

• Starting with a topic, never narrowing to a question. The reader is left guessing what you actually want to find out.
• A question you already know the answer to. If there is no real uncertainty, there is nothing to research.
• A method that does not match the question. Asking why people feel something, then planning only a numbers-based survey, leaves a gap.
• Over-promising. Too many participants, too many methods, too little time.
• Skipping the “so what.” A clear question with no sense of why it matters reads as an exercise rather than research.
• Polishing format before fixing focus. Spacing and citation style matter, but only after the argument underneath is sound.

A simple order of work

If you are unsure where to begin, follow this sequence:

1. Draft a rough question.
2. Read enough to test whether the question is fresh and answerable.
3. Sharpen the question until it names who, what, and how much or how.
4. Write a short justification linking it to a gap.
5. Choose the method that fits, then sketch a timeline.
6. Only now, format and proofread.

A research proposal is really a promise: that you have found a question worth your time, and that you have a sensible plan to answer it. Spend your effort on the question first, and the rest of the document becomes far easier to write.