How to Write a Research Question (Step-by-Step, With Examples)

Almost every stuck research project has the same root problem. Not a weak argument, not thin sources, not bad writing. The research question was never sharp enough to begin with. When the question is vague, everything downstream turns to mush: the literature review has no focus, the methodology has nothing to serve, and the results have no target to hit.

A good research question does the opposite. It tells you which papers to read and which to skip. It tells you what data you need. It tells your marker, in one line, exactly what you set out to find. That is a lot of work for one sentence, which is why it pays to write it carefully.

This guide walks through what a research question actually is, what separates a strong one from a weak one, and a step-by-step process for writing yours. It includes the FINER test, the PICO and PEO frameworks, worked examples across different fields, and the five mistakes that send questions back for revision.

What Is a Research Question?

A research question is the specific, focused question your study sets out to answer. It defines the scope of your project and gives every other part of the paper a job. Your introduction explains why the question matters, your literature review shows what is already known about it, your methodology describes how you will answer it, and your results and discussion report what you found.

Think of it as the anchor of the whole document. A topic is broad, like “social media and mental health.” A research question is narrow and answerable, like “How does daily Instagram use relate to reported anxiety in first-year university students?” The topic could fill a library. The question can be answered in a single study.

Research Question vs Thesis, Hypothesis, and Aim

These four terms get tangled constantly, and mixing them up is one of the fastest ways to confuse a marker. They are related, but they are not the same thing.

The simplest way to keep them straight: the research question asks, the hypothesis predicts, and the thesis answers. Most student papers lead with a research question and one or more sub-questions, and add hypotheses only if the study is quantitative.

What Makes a Research Question Good? The FINER Test

Before you commit to a question, run it through the FINER test, a checklist widely used in health and social science research. A strong question passes all five.

On top of FINER, a good question is usually specific, focused on one issue, and open enough that it cannot be answered with a plain yes or no. “Does exercise help depression?” is a yes-or-no dead end. “How does a twelve-week aerobic exercise program affect depression severity in adults aged 18 to 30?” invites real analysis.

How to Write a Research Question, Step by Step

You do not arrive at a good question by staring at a blank page. You narrow your way to it. Here is the sequence that works for most students.

1. Start with a broad topic you care about

Begin with the general area, not the question. “Renewable energy adoption,” “adolescent literacy,” “small business financing.” At this stage broad is fine, because you are about to narrow it. If you are still at the very start of the paper, our guide on how to start a research paper covers choosing a topic in more depth.

2. Read enough to see what is already known

You cannot find a gap in a field you have not read. Do a focused first pass through recent reviews and key papers to map out what has been established, what is contested, and what keeps getting flagged as “future work.” Our guide on how to find academic sources shows where to look, and how to write a literature review explains how to organise what you find.

3. Find the gap

A research gap is a specific question the existing literature has not answered, or has not answered for your context, population, or time period. Watch for a method that has not been applied to a problem, a group that has been studied in one country but not another, or a finding that two studies disagree on. That gap is where your question lives.

4. Narrow and focus

Take your broad topic and add limits: a population, a setting, a time frame, a variable, an outcome. Each limit you add makes the question more answerable. “Social media and teenagers” becomes “how nightly Instagram use relates to sleep quality in 15 to 17 year olds.” The scope shrinks and the study becomes possible.

5. Draft it with a framework

Now put it into words using one of the frameworks in the next section. Frameworks are not mandatory, but they force you to name the moving parts of your question so nothing important is left vague. Draft two or three versions and pick the sharpest.

6. Pressure-test and refine

Run the draft through the FINER test. Can you answer it with what you have? Is it a real question and not a yes-or-no? Does it hide two questions inside one? Read it aloud to someone who is not in your field. If they can restate what you are trying to find out, the question is clear. If they cannot, keep cutting.

Frameworks That Make It Easier: PICO, PEO, and SPIDER

Frameworks turn a fuzzy idea into a structured question by making you fill in specific slots. Pick the one that fits your kind of study.

PICO (quantitative and clinical)

PICO stands for Population, Intervention, Comparison, and Outcome. It suits studies that test the effect of something.

PEO (qualitative)

PEO stands for Population, Exposure, and Outcome. It fits qualitative work that explores experiences rather than testing an intervention.

SPIDER (qualitative and mixed methods)

SPIDER covers Sample, Phenomenon of Interest, Design, Evaluation, and Research type. It is useful when the design and the type of evidence matter as much as the topic, for example in a synthesis of qualitative studies. The point of any of these frameworks is the same: name the parts, and the question writes itself.

Qualitative, Quantitative, and Mixed Methods Questions

Your research question and your method have to match. In fact, the wording of your question already hints at the method, which is why the two should be written together. The verb you choose usually gives it away.

Quantitative questions

These measure, compare, or test relationships between variables. They tend to start with “how much,” “how many,” “what is the effect of,” or “what is the relationship between.” Example: “What is the effect of class size on exam performance in secondary schools?”

Qualitative questions

These explore meaning, experience, and process. They tend to start with “how,” “why,” or “what does it mean to.” Example: “How do nurses describe moral distress during night shifts?”

Mixed methods questions

These combine both, often as a main question with quantitative and qualitative sub-questions. Example: “How does a peer-mentoring program affect student retention, and how do students experience the support it provides?” Once the question is set, our methodology section guide covers how to design the study that answers it.

Weak vs Strong Research Questions

The difference is almost always specificity. A weak question is broad, vague, or answerable with a single fact. A strong one names the who, the what, and the context. Compare these.

Notice that every strong version adds limits: a population, a setting, a time frame, or a measurable variable. That is the whole move. Keep adding specificity until the question could realistically be answered by one study.

How Many Research Questions Should You Have?

For most projects, one main research question is enough, supported by two to four sub-questions that break it into answerable pieces. The main question captures the overall aim; the sub-questions map onto the sections of your analysis.

If you find yourself with five or six equally weighted main questions, your scope is probably too wide for a single paper. Choose the one that matters most and demote the rest to sub-questions or future work.

5 Common Research Question Mistakes

1. Too broad

“How does poverty affect education?” is a topic, not a question you can answer in one study. Narrow it by population, place, and outcome until it fits your scope.

2. Too narrow or trivial

The opposite problem. If the answer is a single number you could look up, or a fact with no room for analysis, there is nothing to research. Your question should require investigation, not a search engine.

3. A yes-or-no question

“Does homework improve grades?” can be answered in one word, which leaves you nothing to write about. Rephrase it as a “how” or “to what extent” question so the answer has depth.

4. Two questions hiding in one

“How does remote work affect productivity and mental health?” is really two studies. Pick the one that matters most, or split them cleanly into a main question and a sub-question.

5. Not grounded in the literature

A question invented in isolation often turns out to be already answered, or impossible to place in the field. Ground it in what you have read so it clearly extends existing work. This is also where honest research starts: you build on real sources, you read every one, and you can point to exactly where your question came from.

Where Your Research Question Fits

The research question usually appears near the end of your introduction, right after you have established the background and the gap. From there it threads through the whole paper: the literature review justifies it, the methodology answers it, and the discussion returns to it to say whether you answered it and what that means.

If you are mapping out the full shape of your paper, our research paper structure guide shows how every section connects back to the question, and the research proposal guide covers how to frame it before the study begins.