How to Formulate a Research Question, Step by Step

A topic is a subject area, such as sleep, climate migration or nursing burnout. A research question is a single sentence that points a study at one piece of that subject. Most students get stuck because they try to jump straight from the first to the second. Once you know how to formulate a research question in stages, the sentence is much easier to write.

If you want a draft to react to, the free research question generator turns a broad topic into three to five focused questions, each tagged by approach. It is a starting point, so adjust the wording to your class and your instructor's rules.

Topic, narrow, question

Think of a funnel. At the top is your topic. In the middle you narrow it by who, where, when and which angle. At the bottom is a question you can answer with evidence. The first illustration shows the path for a school start time project. If you are still working out how to formulate a research question and have nothing yet, or you are wondering how to come up with a research question at all, begin with the reading step below. A gap in the sources is often where a good question lives.

What a research question does

A good question tells the reader what you will investigate and what kind of evidence will count. It drives your search terms and your method, and it tells you when to stop.

These steps expand the five-step list above. They are the core of how to formulate a research question for any class, so work through them in order and expect to loop back at least once.

Step 1: Start with a topic you can sustain

Choose something that fits the assignment and holds your interest. If your instructor assigned a broad theme, such as public health, pick a corner of it that connects to something you have seen or read.

Step 2: Read before you narrow

Spend an hour in a library database or Google Scholar and skim abstracts. Look for debates and for phrases such as "little is known about" or "further research should." Those phrases mark gaps. Keep a short list of the three most promising ones.

Step 3: Narrow in four directions

Limit the topic by population (who), place (where), time (when) and angle (which aspect). "Sleep and school" becomes "sleep and attendance at two Ohio high schools after a start time change." Each limit cuts the amount of research you have to do, so a longer paper can use fewer limits and a short one needs more.

Step 4: Draft the question as an open sentence

Begin with how, why, what or to what extent. Avoid starting with "is," "does" or "should" unless you plan a clear comparison, because a yes-or-no answer ends the inquiry in one word. A reader should be able to picture the kinds of sources you will use.

Step 5: Test, share and revise

Run the draft through the FINER checklist in the next section, then show it to your instructor, a writing center tutor or a librarian. Expect to revise the wording two or three times.

A widely taught checklist for how to formulate a good research question is FINER, described in Hulley and colleagues' textbook Designing Clinical Research (2007). It was written for clinical research, but the five tests are useful in most fields. Adler University's library page recommends the same criteria for evaluating possible research questions. The table below adapts them for student projects.

Using FINER on a real draft

Take the draft "How does social media affect teens?" It fails F, because no one can survey all teens, and it fails N, because hundreds of studies already cover it. After narrowing, "How do 10th graders at one San Diego high school describe the role of Snapchat in their friendships?" passes F for a short project.

Other quick checks

Many writing centers add three informal tests. Is it arguable, so reasonable people could answer differently? Is it researchable, so evidence exists? Is it specific, so a stranger would know what you mean?

What is the research question in qualitative research asking, in plain terms? It is an open question that asks how or why something happens, or what an experience means to the people living it, and it is answered with words, observations and documents instead of statistics. If you are asking how first-generation students describe their first semester, you are working in this tradition.

The research question generator labels each draft by approach, which can help you decide which tradition fits. Here is how a question for qualitative study projects usually differs from a numeric one, and how to formulate a research question that fits either.

What is the research question in qualitative research made of?

In short, what is the research question in qualitative research made of? Usually three parts: an open verb such as describe or explore, a specific group of people and a setting. The next section looks at each feature.

Features of a question for qualitative study

Qualitative questions tend to use verbs such as describe, explore, understand and experience. They name a specific group and setting. They do not predict an answer in advance, and they allow the data to surprise you. They often begin with "how" because "how" invites a process and a story.

A central question and sub-questions

Many qualitative projects state one central question and two or three sub-questions. If the central question is "How do rural high school students in Montana experience dual-enrollment courses?" the sub-questions might ask about motivation, barriers and support.

Avoid words that imply cause

Avoid "effect," "impact" and "cause" unless your method can support them. Say "describe the role of" or "understand how."

A research question for quantitative research asks about variables, which are things you can measure, and about how they relate or differ. The answer is a number, such as a correlation, a difference between group averages or a percentage. Your wording tells the reader which statistic you will probably use.

Our quantitative research question tool tags drafts in this category and highlights the variables. If you prefer to work by hand, ask which type of research question for quantitative research you need. There are three common ones.

Descriptive questions

A descriptive question asks how much or how often: "What percentage of seniors at one Illinois high school work 15 or more hours a week?"

Comparative questions

A comparative question sets groups side by side: "Do students who use flashcards score differently on Spanish vocabulary quizzes than students who reread notes?"

Relationship questions

A relationship question asks whether two variables move together: "Is there a relationship between weekly hours of sleep and GPA among first-year students?" Say "relationship," not "cause," unless you run an experiment.

Define each variable in your research question for quantitative research

"Stress" is vague. "Score on the 10-item Perceived Stress Scale" is clear. A reader should know what you will count or score before you collect any data.

Research question vs hypothesis is a common point of confusion, and the research question vs hypothesis choice shapes your method section. In the research question vs hypothesis comparison, a research question asks what you want to find out. A hypothesis is a testable prediction of the answer, written before you collect data. The question comes first, and a hypothesis, if you need one, follows from it.

Side by side

Take the question "Is there a relationship between nightly sleep and quiz scores?" The matching hypothesis is "Students who sleep seven or more hours will score higher than students who sleep fewer than six." The question is open. The hypothesis takes a position that data can support or fail to support. The second illustration shows the pair.

The difference between a research question and a hypothesis

The difference between research question and hypothesis comes down to three points. A question can be exploratory, while a hypothesis predicts. A hypothesis needs a variable you can test, while a question may not. And a hypothesis can be supported or not, but a question is simply answered. Experiments, surveys and other quantitative studies usually state both. Qualitative studies and many humanities papers state only the question.

Research question vs hypothesis in a humanities paper

On the research question vs hypothesis divide, history and literature work differently. The counterpart of a hypothesis is usually a thesis statement, which is the claim your paper argues in answer to your question. If you are moving from one to the other, our thesis statement generator can help turn a finished question into a claim. See also what an argumentative essay is.

Many instructors ask for a question only, so check your assignment.

Different courses and journals use different words for the same job, so a list of synonyms for research question can save you confusion. This list of synonyms for research question groups them by how close they are. The terms below are close, but a few are not exact matches, and it helps to know which.

Close synonyms

• Guiding question or driving question: the central question that organizes a project, common in K-12 project-based learning.
• Inquiry question or line of inquiry: the same idea in inquiry-based courses.
• Central question: the main question in qualitative designs, often followed by sub-questions.
• Focus question: a narrower question that frames one section or one lesson.
• Study question: a question the study is built to answer.

Related terms that are not the same

A research problem is the issue or gap that makes the question worth asking. A problem statement explains that problem in a paragraph. A purpose statement says what the study will do. An objective is a specific task that helps answer the question. Treat these as partners, not substitutes.

Research question statement and synonyms for research question in a proposal

A research question statement is just the question written as a sentence in your introduction, usually after the gap. Readers expect to find it in the last lines of the introduction, as in the fourth illustration. Some instructors call this the statement of research question, so use the term on your assignment sheet, and check it against the synonyms for research question above. For the paragraph that leads up to it, see our guide to how to write an introduction.

Nursing, medicine and other health programs often use PICO to formulate a question. PICO stands for Patient or Population (sometimes Problem), Intervention, Comparison and Outcome, and PICOT adds Time, the period over which the outcome is measured. A clinical question fills those slots: in adults with type 2 diabetes, does a weekly telehealth call, compared with usual care, change A1C levels over six months?

You do not need PICO for a humanities paper. Use it when the topic is a treatment or nursing practice and your instructor asks for evidence-based practice. The PICOT question generator assembles the parts into one sentence and suggests a search for PubMed.

The International Baccalaureate Diploma Programme asks every student to write an extended essay, an independent research paper of up to 4,000 words. The essay is built around a focused question, so the extended essay research question carries real weight, and everything on this page about how to formulate a research question applies.

Narrow the topic until the question fits the word limit, and check that sources exist before you commit. The IB has introduced a new extended essay guide, with first assessment in 2027, so take the current wording of the requirements and criteria from your coordinator and supervisor rather than from older handouts.

Most weak attempts at how to formulate a research question break one of a handful of rules. Check your draft against this list, and use the paired samples earlier on the page as a reference.

Too broad or too narrow

If you could write a book on it, the question is too broad. If one source answers it, it is too narrow. Adjust by adding or removing a limit on population, place or time.

Built-in answers

A question such as "Why is online school worse?" assumes the conclusion. Rewrite it neutrally so evidence can point either way.

Writing a topic, not a question

"Climate change in Florida" is a topic. A question has a verb and a focus, and it ends with a question mark.

Changing the question too late

Refine the question as you read, but settle it by the first third of the project. Changing it the night before a deadline means redoing the research.

When you are ready for a draft or fresh options, the research question examples tool shows tagged samples for your own topic, and you can browse more student writing guides for the next step in your paper.