A research question is the sentence your whole project answers. It shapes the literature you read, the method you choose, the data you collect and the conclusions you can reach. When a dissertation or research essay goes wrong, the cause can often be traced back to a question that was too broad, too vague or impossible to answer with the time and resources available.
This guide explains what makes a research question strong, sets out a practical process for turning a topic into a question, and shows weak questions alongside improved versions so you can see the difference.
Why the research question matters
Examiners read your work asking one thing above all: does it answer the question it set? A strong question gives your project a clear purpose and makes many later decisions easier. It tells you which sources are relevant, which method fits, and when you have enough evidence. A weak question leaves every decision open, and the result is often a project that describes a topic without answering anything.
What makes a research question strong
A strong research question is usually:
- Focused: it addresses one clear issue rather than a whole field.
- Researchable: it can be answered with evidence you can actually obtain.
- Clear: its key terms are specific and can be defined.
- Significant: answering it would add something useful, even if modest, to understanding or practice.
- Feasible: it fits your word limit, your deadline, your skills and any ethical constraints.
- Open: it invites analysis rather than a simple yes or no, unless your design is set up to test a specific hypothesis.
From topic to question: a practical process
- Start with a broad topic that genuinely interests you, because you will spend months with it.
- Read background material to find out what is already known and what is debated. Review articles are especially useful at this stage. Our guide to reading journal articles efficiently will help you move quickly.
- Narrow the topic by choosing a population, a place, a time period, or one aspect of the issue.
- Identify a problem or puzzle. Look for a disagreement between studies, a group that has not been studied, a recent change that existing research does not cover, or a claim that has not been tested.
- Write the question down, then check what type of question it is, because different types need different methods.
- Test it against the criteria above, and discuss it with your supervisor or tutor before you commit.
Expect to revise your question several times. Refining it as you learn more is a normal part of research.
Types of research question
- Descriptive: what is happening? For example, how do students use library study spaces during the examination period?
- Comparative: how do two or more groups, places or periods differ?
- Relational: is there a relationship between two variables?
- Causal: does one factor influence another? Causal questions need designs that can support causal conclusions, such as experiments or carefully controlled comparisons.
- Exploratory or interpretive: how do people experience or make sense of something? These questions often suit qualitative methods such as interviews.
- Evaluative: how effective is a policy, programme or intervention, and by what measure?
Weak questions and improved versions
Hypothetical examples: these questions are illustrations written for this guide. They do not describe real studies.
Weak: What is the effect of social media?
Improved: How do first year undergraduates describe the role of social media in their sense of belonging during their first term at university?
Why it is better: it names a population, a period and a specific outcome, and it suits interviews or focus groups.
Weak: Is remote work good?
Improved: Is there a relationship between the number of remote working days per week and reported job satisfaction among employees of small firms in Dubai?
Why it is better: "good" has been replaced by a measurable outcome, and the scope is limited to a defined group.
Weak: Why is renewable energy important?
Improved: What factors do policy documents and installers identify as influencing household adoption of rooftop solar panels in Abu Dhabi since 2015?
Why it is better: it can be answered with specific evidence, it does not assume its own answer, and the time frame is clear.
Weak: Does assessment stress students?
Improved: How do final year nursing students perceive the effect of clinical placement assessments on their wellbeing?
Why it is better: it moves from a yes or no question to an open one about a defined group and context.
Using a framework
Some disciplines use frameworks to structure questions. In health sciences, PICO is widely used for clinical and evidence based practice questions: Population, Intervention, Comparison and Outcome. For example: in adults with type 2 diabetes (population), does a structured exercise programme (intervention), compared with usual care (comparison), improve blood glucose control (outcome)? Use a framework if your department recommends one, but remember that it organises a question; it cannot make a poor question good.
Sub questions, aims and objectives
Larger projects often break the main question into two to four sub questions, or express it as an aim supported by objectives. Each sub question should contribute to answering the main question, and together they should be achievable within your project. If a sub question would need a separate study to answer, it probably belongs to a different project.
Test your question
- Can I state my question in one clear sentence?
- Have I defined, or can I define, every key term?
- Do I know what evidence would answer it, and can I obtain that evidence?
- Can I answer it within my word limit and deadline?
- Does it avoid assuming the answer?
- Does my literature review show why it is worth asking?
- Has my supervisor or tutor seen it?
Common mistakes and how to avoid them
- Too broad: narrow the population, place, time or aspect.
- Too narrow or trivial: make sure the answer would matter beyond a single fact.
- Several questions in one: split them, or choose one as the main question.
- Built in assumptions: "Why does X cause Y?" assumes that it does. Ask whether, and how.
- Vague terms: replace words such as "good", "effective" or "impact" with something you can define and measure or explore.
- Unavailable data: check early that you can access the people, documents or datasets you need, including any ethical approval required.
Frequently asked questions
What is the difference between a research question and a hypothesis?
A research question asks what you want to find out. A hypothesis is a testable prediction, usually used in quantitative research, stating the answer you expect. Many qualitative projects use research questions without hypotheses.
Can my research question change during the project?
Yes, it often becomes sharper as you read and collect data. Agree any significant change with your supervisor, and make sure your final write up matches the question you actually answered.
How do I know if my question is original enough?
Expectations depend on your level of study and your programme. At undergraduate level, applying an established question to a new group, place or period is often acceptable; postgraduate research is usually expected to make a clearer contribution. Check your handbook and ask your supervisor. Your literature review should show where your question fits.
Related resources: literature review checklist, dissertation timeline and planning guide, and our Research Finder.
Related articles on our blog: Writing Strong Research Questions: Weak and Strong Examples, How to Write a Dissertation Proposal and How to Structure a Literature Review.
References
None. This article gives general guidance and does not cite specific sources.
Further reading
Bell, J. and Waters, S. (2018) Doing Your Research Project: A Guide for First Time Researchers. 7th edn. Open University Press.
Booth, W.C., Colomb, G.G., Williams, J.M., Bizup, J. and FitzGerald, W.T. (2024) The Craft of Research. 5th edn. University of Chicago Press.
Note: the published title of Bell and Waters prints "First Time" with a hyphen. It is written without one here in line with this site's house style.
