How to Develop a Clear and Researchable Question

A research question turns a broad interest into a problem that can be investigated with evidence. It gives the search, method and final argument a common direction. A good question is not simply impressive wording; it must fit the task, discipline, access and time available.

Questions often change during preliminary reading. That is normal when the change is recorded and the project remains aligned. For dissertations or studies involving people, major changes may also require supervisor or ethics approval before data collection.

Start with a real area of curiosity

Write what you want to understand and why it matters in the course or field. Avoid choosing a topic only because it sounds important. Sustained interest helps when the initial idea must be narrowed and revised.

Map what the task allows

Record the word count, timeframe, evidence requirements, geographical limits and permitted methods. A question suitable for a doctoral project may be impossible in a six-week module, even when it is intellectually strong.

Do a preliminary source scan

Read a small set of current and foundational sources to learn the terminology, active debates and available evidence. The aim is orientation, not a hidden literature review completed before the question exists.

Turn the topic into a problem

Ask what is uncertain, changing, disputed, poorly explained or important to compare. “Artificial intelligence in education” is a topic; a research question specifies a relationship, experience, process or interpretation that can be examined.

Specify the key elements

Identify the population or material, phenomenon, context, place and period where relevant. Include only details that genuinely define the inquiry. Overloading the sentence can make the question precise in appearance but incoherent in practice.

Choose an appropriate question form

Descriptive, comparative, explanatory, evaluative and interpretive questions require different evidence. “What”, “how” and “why” are starting points, not guarantees. Make sure the grammatical form matches the intellectual task.

Avoid a predetermined answer

Remove leading language that assumes harm, success or causation before the research. A question may be motivated by a concern while still allowing the evidence to challenge the expected answer.

Test whether it is answerable

List the data or sources needed and the method that could produce a defensible answer. If the necessary records, participants, languages or equipment are inaccessible, revise the question before promising the study.

Check ethical and practical access

Research involving people, sensitive topics or private records may need formal approval and secure handling. Do not contact participants or collect data until the relevant institutional process is complete.

Match scope to available depth

Narrow one dimension at a time: population, place, period, concept or outcome. Stop when the question can be answered with sufficient analysis, not when it becomes so narrow that no meaningful evidence exists.

Define ambiguous terms

Write an operational or conceptual definition for important words such as success, access, engagement or quality. If several plausible definitions would change the study, the question needs clarification or an explicit choice.

Check contribution without exaggeration

State what the answer may clarify for this context. An undergraduate project does not need to solve an entire field. A transparent replication, comparison or local application can be worthwhile when framed proportionately.

Write a working question and alternatives

Keep two or three versions and compare what each would require. Discuss the strongest option with the supervisor or tutor. Record why the final wording was selected and what it excludes.

Recheck alignment before data collection

Confirm that the literature review, method, variables or interview prompts and planned analysis all address the same question. If they do not, revise the design while changes remain possible and approved.

Separate the question from the expected answer

A working hypothesis, proposition or thesis may express an expected relationship, while the research question states what will be investigated. Keep the distinction clear so evidence can alter the conclusion. In exploratory or interpretive work, a formal hypothesis may be inappropriate; use the terminology required by the discipline instead of adding one for appearance.

Keep a record of narrowing decisions

Save the broad topic, search terms, rejected versions and reasons for the final scope. This record helps explain the project to a supervisor and prevents the question from expanding silently as new sources appear. It also makes later limitations easier to state because the reader can see which populations, periods or outcomes were intentionally outside the inquiry.

Use supervisor feedback to test feasibility

Bring the working question, likely evidence, method, access constraints and a short list of exclusions to the discussion. A question can sound clear while requiring data or expertise the project cannot obtain. Record the agreed wording and any approval condition so later revisions do not quietly change the study’s scope.

Final checklist

• The question is clear without a paragraph of explanation.
• Its scope fits the assignment, time, access and available evidence.
• The wording does not assume the conclusion.
• A suitable and ethical method could answer it.
• Every major project section remains aligned with the same question.

Browse programmes listed on Univs, compare participating institutes and see how the application journey works. For an assessed project, confirm the question with the relevant tutor, supervisor and ethics process.