Narrow a dissertation question to something you can investigate

Move from a subject to a question

Start with the broad area, then identify the uncertainty that interests you. 'Urban transport' is a subject; a question asks about a relationship, explanation, interpretation or comparison within it. Write several candidate questions rather than polishing the first title that comes to mind.

For each candidate, name the main concepts and explain what would count as an answer. Avoid questions whose wording assumes the conclusion. If you already describe a policy as successful, ask whether that judgment is what you intend to investigate rather than a premise the project has established.

Set explicit boundaries

Choose relevant limits such as a time period, location, case, text collection or population. Explain why those limits make sense for the question. A boundary should support an interpretable study, not merely shrink the material until it fits a convenient folder.

Check whether the remaining question still has enough substance for the assignment. Narrowing by removing all comparison or uncertainty can leave only description. Use a short statement of inclusion and exclusion: what the project examines, what it does not examine and what limitation follows from that choice.

Test access to evidence

Locate a small sample of the material you would need. Confirm whether it is available, readable and usable under the relevant conditions. A question depending on records you cannot obtain or participants you cannot recruit may need revision before a detailed plan is worthwhile.

Ask the supervisor about approval requirements before contacting participants or collecting data. Use the university's process for research ethics and access arrangements. Do not treat a public-looking source or an informal conversation as automatically suitable research data. Record unresolved access issues as project risks that need decisions.

Match the method to the question

Describe how you would move from the available material to an answer. If you propose a comparison, identify the basis of comparison. If you propose analysis of texts or data, explain what you would examine and what the method can establish. Avoid choosing software first and inventing a question to justify using it.

Check the skills and time the method requires. You may need training, a pilot task or a simpler design. Discuss those needs honestly with the supervisor. A familiar method used carefully can support a stronger feasible project than an ambitious method you cannot implement or explain within the course.

Take options to supervision

Prepare a short comparison of the strongest candidate questions, their evidence, boundaries and main difficulty. Ask the supervisor which option fits the assessment and available support. Keep a note of the reasons for the agreed direction rather than recording only the final title.

After the discussion, write a provisional question and a first research task that tests it. Revisit scope when the initial evidence reveals something unexpected. Changes should be deliberate and discussed where required; avoid letting the project expand each time you encounter an interesting source that answers a different question.

Your next steps

• Turn the broad subject into a question with an answerable uncertainty.
• State and justify the project's boundaries.
• Test access and approval needs before collecting material.
• Compare feasible options with your supervisor.

Keep planning

• Prepare a dissertation update your supervisor can act on
• Prepare for presentation questions without scripting every answer
• Choose a Model Scale for Your Design Question
• Find note-taking support that works for your course