A PhD topic has to satisfy two very different tests. It must be intellectually significant enough to contribute new knowledge, and practical enough that one person can complete it with the time, data and resources actually available. Topics fail far more often on the second test than the first.

Start from the literature, not from enthusiasm

Enthusiasm matters, but it is not a topic. Begin by reading recent review articles and the “future research” sections of recent theses and papers in your area. Researchers routinely state exactly what they could not address.

• Read systematic reviews to see where evidence is thin or contested
• Check the limitations sections of key papers
• Look at conference proceedings for emerging questions
• Note where two respected studies reach conflicting conclusions

Types of legitimate research gap

Note that “nobody has studied X” is sometimes a warning rather than an opportunity. Ask why not – often the data is inaccessible or the question is not meaningful.

The feasibility test

Before committing, answer honestly:

1. Can I access the data or participants I need, realistically?
2. Will I get ethical approval, and how long will that take?
3. Do I have or can I learn the required methods and tools?
4. Can this be completed in three to four years alongside teaching duties?
5. Is funding available for any equipment, travel or transcription?
6. Is there a supervisor here with genuine expertise in this area?

A brilliant question with inaccessible data is not a viable PhD. Scope down until every answer is yes.

Narrow it properly

Doctoral topics almost always start too broad. Narrow along several dimensions at once:

Each workable version specifies population, context, variables and boundary. That specificity is what makes a study completable and defensible.

Match the topic to supervision

Supervision quality strongly predicts completion. Before finalising, identify supervisors whose published work genuinely overlaps your area, read several of their recent papers, and approach them with a short, specific proposal rather than a vague interest.

A supervisor with adjacent but not matching expertise can work, but you will carry more of the methodological learning alone.

Warning signs of a topic that will not survive

• You cannot state the question in one clear sentence
• Data access depends on an organisation that has not yet agreed
• The scope keeps expanding every time you read more
• No supervisor in your department works anywhere near it
• It requires a method you would need years to learn from scratch
• The contribution is unclear even to you

It is far cheaper to abandon a topic in month two than in year two.