How do I determine my dissertation topic?

A dissertation topic must do more than hold your attention. It must support an important question, a credible design, and a project that can be completed with the time, data, and guidance you actually have.

Separate the topic from the question

“Audit quality,” “tax avoidance,” or “employee disclosure” is a topic area, not a dissertation. A research question names a relationship, mechanism, decision, or measurement problem within that area. Moving from topic to question is the first real act of scoping.

Write the question in one sentence. Then explain who could change a belief or decision after seeing the answer. If that second sentence is vague, the project is not ready.

Three tests a candidate must pass

1. Contribution: Would a credible answer change how researchers understand the setting, mechanism, or method?
2. Feasibility: Can you obtain the evidence, permissions, skills, and advising needed to answer it?
3. Durability: Will the question still matter after the dissertation, and can it lead to related work?

Your personal interest matters too, but fascination cannot rescue an infeasible design. Conversely, a convenient dataset does not create a contribution by itself.

Generate candidates before choosing

Maintain a question log while reading papers, attending workshops, working with faculty, and learning institutional details. Record the puzzle, why it matters, possible evidence, and the largest threat. Review the log periodically with an adviser rather than falling in love with the first workable idea.

Replication, extension, and new-setting ideas can all be useful starting points. The key is to identify what the new evidence would teach, not merely what feature differs from prior work.

Scope the smallest convincing study

A dissertation should demonstrate independent research judgment, not solve an entire literature. Narrow the population, mechanism, outcome, or claim until one design can answer the question cleanly. A focused project that survives scrutiny is more valuable than a sweeping project that never stabilizes.

Use the committee early

Ask potential committee members what would make the project fail, what literature you are missing, and whether the design fits the claim. Resolve authorship, data ownership, and independence expectations before a collaborative project becomes the dissertation.

A good chair helps you make decisions, not avoid them. Bring alternatives and your recommendation to meetings so feedback can focus on judgment rather than status updates.

Warning signs

• The project depends on access or permission that no one has confirmed.
• Its importance rests mainly on being the first use of a method or dataset.
• The central hypothesis changes whenever a result changes.
• Only one person can advise the design or evaluate the setting.
• You cannot name a smaller version that would still teach something.

Topics evolve during a dissertation. That is normal. The goal is not to predict every revision; it is to choose a question strong enough to improve through them.