A research question is the question your study sets out to answer. A good one is focused on a single issue, researchable with the data and time you have, specific about who, where and when, and complex enough that it cannot be answered with yes or no. To write one, choose a topic, find what is not yet known, narrow the scope, choose the type of question and check it against the FINER criteria. This guide shows each step with examples.
Why the research question matters
Your research question decides almost everything else in your study: which literature you review, which method you use, what data you collect and what your conclusion must answer. A vague question leads to a vague thesis; a precise question makes every later decision easier.
How to write a research question in 5 steps
1. Choose a broad topic
Start with an area that interests you and matters in your field, for example: social media and mental health.
2. Find what is not yet known
Read recent studies and reviews. Look at their "limitations" and "future research" sections, and notice where findings disagree. These are your research gaps. See how to find a research gap.
3. Narrow it down
Decide who (population), where (setting) and when (period), and which aspect of the topic you will study.
- Broad: How does social media affect mental health?
- Narrower: How does time spent on short-video apps relate to sleep quality among first-year undergraduates?
4. Choose the type of question
The type of question tells you what kind of study you need.
| Type | What it asks | Example |
|---|---|---|
| Descriptive | What is the state or extent of something? | What proportion of first-year undergraduates report poor sleep quality? |
| Comparative | How do groups or settings differ? | How does sleep quality differ between students living on and off campus? |
| Relational | How are two variables associated? | What is the relationship between daily short-video use and sleep quality among first-year undergraduates? |
| Causal | Does one thing affect another? | Does a two-week limit on evening phone use improve students' sleep quality? |
| Exploratory | How is something experienced or understood? | How do first-year students describe the role of their phones in their bedtime routines? |
| Explanatory | Why does something happen? | Why do some students keep using their phones late at night despite poor sleep? |
| Evaluative | How well does something work? | How effective is a university sleep-hygiene workshop in improving students' sleep? |
Quantitative studies usually ask descriptive, comparative, relational or causal questions. Qualitative studies usually ask exploratory or explanatory "how" and "why" questions. See qualitative vs quantitative research.
5. Check it with FINER
The FINER criteria are a widely used checklist for research questions:
- Feasible: can you answer it with the time, data, skills and access you have?
- Interesting: does it interest you, your supervisor and your field?
- Novel: does it add something new, even if only a new setting or population?
- Ethical: can it be studied without harming participants?
- Relevant: does the answer matter to research, practice or policy?
For clinical and health questions, the PICO framework helps you make the question specific: Population, Intervention, Comparison, Outcome.
Among adults with type 2 diabetes (P), does a text-message reminder program (I), compared with usual care (C), improve medication adherence over six months (O)?
Weak vs strong research questions
| Weak | Problem | Stronger |
|---|---|---|
| Is social media bad for students? | Yes/no, vague | How does daily time on image-based social media relate to body satisfaction among female undergraduates aged 18–21? |
| What is climate change? | Answered by a textbook | How have rising summer temperatures affected wheat yields in northern India since 2000? |
| Why do companies fail? | Far too broad | Why do family-owned retail businesses in rural Ireland struggle to adopt online sales? |
Main question and sub-questions
A thesis often has one main research question and two to four sub-questions that break it into parts. Each sub-question is usually answered in a chapter or a section of the results.
Main question: How do first-year undergraduates use their phones at night, and how is this related to their sleep quality?
- How much time do students spend on their phones after 10 pm?
- Which apps account for most of this time?
- How is evening phone use related to sleep quality scores?
- How do students explain their late-night phone use?
From research question to hypothesis
If your question asks about a relationship or an effect, you can turn it into a testable prediction. "Is daily short-video use related to sleep quality?" becomes "Students who spend more time on short-video apps after 10 pm report poorer sleep quality." See how to write a hypothesis.
Try the free research question generator
The research question generator suggests research questions for your topic, labels each one by type and approach, and explains what makes it researchable. Add your population, setting or period to get more specific questions.
Frequently asked questions
How many research questions should a thesis have?
Usually one main question with two to four sub-questions. Too many questions spread a thesis too thin.
Can a research question be a yes or no question?
It is better to avoid it. "Is X related to Y?" becomes stronger as "How strongly is X related to Y, and for whom?"
Where does the research question go in a thesis?
In the introduction, after you explain the problem and its background. It is repeated or referred to in the methodology, results and conclusion.
Can my research question change?
Yes, especially early on. Refine it as you read the literature, but agree any major change with your supervisor.