To write a research question, narrow a broad topic to one specific, answerable question that your project can address with evidence. A good question decides what you read, which method you use and what your conclusion must deliver.
This guide covers the qualities of a strong question, a step-by-step method, useful frameworks, types of question for different designs, and before-and-after examples from several subjects.
The Five Tests of a Strong Research Question
A strong research question is focused, researchable with available sources or data, feasible in your time frame, complex enough to need analysis rather than a yes-or-no fact, and relevant to a real debate or problem in your field.
If your question fails one test, adjust its population, setting, variables or time frame until it passes.
What a Research Question Is and Does
A research question is the single question your paper, dissertation or project sets out to answer. It sits at the end of your introduction, after the background and the gap in existing research.
It works as a filter. Sources that help answer it belong in your literature review; sources that do not, however interesting, stay out. The method must be able to answer it, and the conclusion must answer it directly.
Many projects add two or three sub-questions that break the main question into parts. Each sub-question often becomes a section of the results or discussion.
Main question: How has fully remote work affected staff retention in mid-sized UK software firms since 2021?
Sub-questions: 1. How have voluntary turnover rates changed since the move to remote work? 2. Which reasons do departing staff give in exit data? 3. Which retention practices do firms with lower turnover report using?
Notice that each sub-question can be answered with a specific kind of evidence, and that together they answer the main question. If a sub-question does not contribute to the main one, cut it.
Qualities of a Good Research Question
| Quality | What it means | Warning sign |
|---|---|---|
| Focused | Names a specific population, setting, variable or text | Could fill a whole book |
| Researchable | Evidence or data exists or can be collected | Depends on information nobody can access |
| Feasible | Answerable within your time, word count and resources | Needs years of data or a large funded team |
| Complex | Needs analysis, comparison or interpretation | Answered by a single fact or a search engine |
| Relevant | Connects to a current debate, gap or practical problem | Nobody would act differently whatever the answer |
| Neutral | Does not assume the answer in its wording | "Why is X harmful?" before showing that it is |
Writing Your Question Step by Step
- Choose a broad topic that interests you and fits the brief, such as social media and mental health.
- Do preliminary reading. Skim recent reviews and articles to see what is known and what is debated.
- Find a gap or problem. Look for limitations sections, conflicting findings and calls for further research.
- Narrow the focus. Choose a population, setting, time period, variable or text.
- Draft the question using a question word that matches your aim: how, what, to what extent, why.
- Test it against the five qualities and check that suitable sources or data exist.
- Refine the wording until every key term is precise and the question does not assume its own answer.
Topic: social media and mental health.
Narrowed: passive scrolling, loneliness, first-year university students.
Question: How is time spent passively scrolling social media related to loneliness among first-year university students living away from home?
Types of Research Question
The type of question you ask commits you to a type of design. Choose deliberately.
| Type | Asks | Example | Typical method |
|---|---|---|---|
| Descriptive | What is happening, and how much? | What proportion of small firms in Nairobi use mobile payments for supplier invoices? | Survey, secondary data |
| Comparative | How do groups differ? | How do remote and office-based graduates differ in promotion rates after three years? | Comparison of groups |
| Relational | How are variables connected? | Is sleep duration associated with exam performance in sixth-form students? | Correlational study |
| Causal | Does X affect Y? | Does a peer mentoring scheme reduce first-year dropout? | Experiment or quasi-experiment |
| Exploratory | How do people experience or understand X? | How do newly qualified nurses experience their first night shifts? | Interviews, focus groups |
| Evaluative or interpretive | How well, or in what way, does X work or mean? | To what extent does the novel challenge Victorian ideas of class? | Close reading, critical analysis |
Frameworks That Help You Structure a Question
Frameworks act as checklists that make sure your question names every important element.
- PICO or PICOT (health sciences): Population, Intervention, Comparison, Outcome, Time.
- SPIDER (qualitative health research): Sample, Phenomenon of Interest, Design, Evaluation, Research type.
- PEO (exploratory studies): Population, Exposure, Outcome.
- FINER (a quality check): Feasible, Interesting, Novel, Ethical, Relevant.
PICO example: In adults with type 2 diabetes (P), does a nurse-led text message reminder programme (I), compared with usual care (C), improve medication adherence (O) over six months (T)?
Outside health subjects you rarely need a named framework, but the same logic applies: who or what, which factor, compared with what, and measured or interpreted how.
Before and After: Examples by Subject
| Subject | Weak question | Strong question |
|---|---|---|
| Business | Is remote work good for companies? | How has fully remote work affected staff retention in mid-sized UK software firms since 2021? |
| Nursing | How can falls be prevented? | Does purposeful hourly rounding reduce falls among older inpatients on medical wards compared with usual care? |
| Education | Does technology help students learn? | To what extent does weekly low-stakes online quizzing improve end-of-term maths results in Year 9 pupils? |
| Psychology | Why do people procrastinate? | How is fear of failure related to academic procrastination among undergraduate students during exam periods? |
| History | What caused the First World War? | How did British press coverage shape public support for intervention in the summer of 1914? |
| Literature | What is the novel about? | How does the narrator's unreliability in the novel complicate its critique of memory and guilt? |
Each strong version names a specific group, setting or text, a clear focus and, where relevant, a time frame. None can be answered with a quick fact.
Research Questions, Hypotheses and Aims
Quantitative studies often turn the research question into a hypothesis: a testable prediction of the answer. For example, "Students who take weekly quizzes will score higher on the end-of-term test than students who do not."
Qualitative studies usually keep an open question without a hypothesis, because they explore rather than test. Many proposals also state an aim, the overall purpose in one sentence, and two to four objectives that describe the steps to answer the question.
Keep all of these aligned. The aim, question, hypothesis and objectives should describe the same project in different forms. When you revise one, check the others in the same sitting so they never drift apart.
Common Mistakes to Avoid
- Too broad: "What are the effects of climate change?" could be a library, not a paper.
- Too narrow or factual: "When was the policy introduced?" has a one-line answer.
- Loaded wording: "Why is social media destroying attention spans?" assumes the answer.
- Vague terms: words such as "effective", "better" or "young people" left undefined.
- Mismatched method: a causal question answered with a one-off survey.
- Unanswerable in time: a question that needs data you cannot reach before the deadline.
How EssayHorse Helps with Research Questions
Send your topic, the brief, your discipline and level, and any reading or tutor feedback so far. A Standard, Master or Elite writer can prepare a custom research proposal or paper built around a focused question, with background, gap, method and aligned objectives.
Your work is written from scratch and plagiarism-checked, for study and reference. Deadlines run from 3 hours to 20 days, the price is shown before you pay, and free revisions within the original scope are included for 14 days. We never invent data or findings.
Want a Project Built on a Sharp Question?
Send your topic, brief and any reading so far. Your writer prepares a custom proposal or paper built around a focused, answerable research question.
Order Your Research ProposalFree revisions within scope for 14 days · Full refund if late · Written from scratch for your order
Frequently Asked Questions
Usually one sentence of 15 to 35 words. It should be long enough to be specific and short enough to remember.
Yes. Many have one main question and two or three sub-questions that break it into parts.
A research question asks what the project will investigate. A thesis statement gives the answer or position the paper will argue.
Usually at the end of the introduction, after the background and the gap in existing research.
Generally no. Questions starting with how, what, why or to what extent invite the analysis tutors expect. Causal questions such as "does X reduce Y" are the main exception.
Quantitative studies usually include one or more hypotheses. Qualitative studies normally do not.
A health sciences format naming the Population, Intervention, Comparison and Outcome, with Time added in PICOT.
Small refinements are normal as you read more. Major changes should be discussed with your supervisor and reflected in every section.