How to Write a Research Question (With Examples)

Every thesis stands or falls on its research question. A good one tells you exactly what data to collect, what literature is relevant, and when you are finished. A bad one makes every chapter harder: the literature review sprawls, the methods don't quite fit, and the conclusion answers something slightly different from what was asked. The encouraging news is that writing a strong research question is a craft with clear rules, and weak questions can almost always be repaired.

What Makes a Research Question Good

A strong research question is:

1. Focused. It targets one specific problem, not a whole field.
2. Answerable. You can address it with evidence you can realistically collect in your timeframe, with your access and resources.
3. Open. It cannot be settled with "yes," "no," or a quick Google search. Questions starting with how, why, to what extent, or what explains usually do more analytical work than ones starting with is or does alone.
4. Arguable in its answer. The answer requires interpretation of evidence, not recitation of facts.
5. Significant. Someone in your field should care about the answer, which usually means it connects to a gap or debate in the literature.

Tip: A quick self-test — if you can imagine two competent researchers reaching different, defensible answers to your question, it is probably a real research question. If everyone would answer identically in one sentence, it is a lookup, not a question.

Good vs Bad: Before and After

The fastest way to internalize the rules is to watch weak questions get repaired.

Notice the repeated repair moves: a specific population replaces "people," a specific place and period replace "the world," and an analytical verb (how, why, what explains) replaces a yes/no or value-laden framing.

Four Techniques for Narrowing a Broad Question

Most students start too broad, which is normal. Narrow deliberately along one or more of these dimensions:

• Population. Not "students" but "first-generation undergraduate students"; not "companies" but "family-owned manufacturing SMEs."
• Place and time. Not "worldwide" but "in Ontario"; not "throughout history" but "between 2015 and 2024."
• Aspect or mechanism. Not "the effects of remote work" but "the effect of remote work on informal mentoring." Choose one variable, theme, or causal pathway.
• Comparison. Adding a contrast often sharpens everything: "Why did policy X succeed in country A but fail in country B?" gives you built-in analytical structure.

Apply these one at a time and stop when the question matches what you can actually study. A question can also be too narrow: if only three documents or five participants on Earth are relevant, widen one dimension.

Research Question vs Hypothesis

These are related but not interchangeable, and many theses need both.

The workflow runs in one direction: the question comes first, and the hypothesis, if you have one, is your predicted answer to it, derived from theory or prior findings. Exploratory and qualitative studies often proceed on questions alone, and that is methodologically legitimate; forcing a hypothesis onto an exploratory design just to look "scientific" weakens both.

A Final Checklist Before You Commit

Run your question through these checks, ideally with your supervisor:

• Can I name the exact data or sources that would answer this?
• Can I collect that evidence with my time, budget, and access?
• Does the question connect to a gap or debate I can cite in my literature review?
• Is every term in the question defined or definable? ("Effectiveness," "success," and "impact" all need operational definitions.)
• Does it still interest me enough to live with for a year or more?

Expect your question to evolve as you read and gather data; refining it is progress, not failure. Just make sure the final thesis and the final question match. For more worked examples, the Purdue Online Writing Lab and the UNC Writing Center both have free guides on developing research questions.

Frequently Asked Questions

How many research questions should a thesis have?
Usually one central question, sometimes supported by two or three sub-questions that break it into parts. More than that and the project tends to lose focus — each question multiplies the data and analysis you owe the reader.

What is the difference between a research question and a thesis statement?
The question is what you set out to answer; the thesis statement is the answer you arrive at and argue for. You draft the question first and refine the statement as your evidence comes in. See how to write a thesis statement for the second half.

Can a research question change during the thesis?
Yes, and it often should. Reading and early data routinely reveal that a question was too broad or aimed at the wrong thing. The only rule is that your finished thesis must answer the question it finally states.

Related Guides

• How to write a thesis statement — turning the question into an argument
• How to write a literature review — where the gap behind your question comes from
• How to write a thesis — how the question fits the whole project

Once your question is set, Grongy helps you build the thesis around it: your drafts stay grounded in the sources you upload, and your citations stay consistent as the argument grows, so every chapter keeps answering the question you actually asked. The question, and its answer, are yours.