Beginner's Guide to Selecting the Perfect PhD Topic

TL;DR

Students should not let a supervisor’s current interests automatically determine the PhD topic; the student should lead the search for a suitable research question.

Briefing Cornell Notes

Briefing

A strong PhD topic doesn’t come from a supervisor’s preferences—it comes from the student’s own search for something genuinely novel, relevant, and doable within the real constraints of a multi-year degree. Supervisors often steer toward work that matches their current research, which can leave students with projects that are either too narrow to matter or too broad to finish. The practical fix is to choose the research question yourself, then align it with your supervisor.

Novelty is the first filter. The goal isn’t to “change the world,” but to find a contribution that’s at least a little new—something that hasn’t been done before, or a fresh angle on an existing question. That requires reading the literature and actively probing for gaps: try slightly larger or smaller versions of a question, test whether a “tiny” new sub-question has already been answered, and keep checking until the novelty becomes clear. To speed up discovery, the transcript recommends using semantic search tools (for example, Elicit) by entering a research question and scanning what comes up, then repeating the process with other search systems (such as Consensus) to see whether the same themes appear and what remains unresolved. Over time, repeated literature review builds a “sixth sense” for what’s genuinely new.

Relevance comes next—because a research question should have clear beneficiaries. One way to gauge what matters now is to read science journalism outlets like ScienceAlert or similar “hot topic” coverage. Searching for a broad term (like bats) can surface current public-facing concerns and emerging storylines—such as links between bats and pandemics—that can help students judge whether their topic will attract attention beyond academia. The underlying logic is pragmatic: funding and institutional support tend to follow areas that look timely and important, so aligning with a “hot” theme can improve the odds of long-term research value.

Feasibility is the third and arguably most decisive criterion. A topic must fit the timeline of a PhD—often about two years for the core work—because PhDs involve repeated failure and iteration. The transcript urges students to ask what can be accomplished in 2–3 years, not 5–10, and to plan around access to data, equipment, expertise, ethics approvals, and paperwork. Scope should be tuned through trial: too narrow can trap a student when experiments fail; too broad can become unfinishable. The “sweet spot” sits in the middle, with enough flexibility to pivot when results don’t go as expected.

Finally, the best projects aren’t invented in isolation. Students should pressure-test ideas by talking with supervisors, peers, other PhD students, and postdocs in the target research group. The strongest topics are refined through multiple rounds of feedback before they become the foundation for a thesis or dissertation.

Cornell Notes

A good PhD topic balances novelty, relevance, and feasibility—rather than simply matching a supervisor’s interests. Novelty means finding a small but real research gap by repeatedly reading and testing variations of questions, using tools like semantic search (Elicit) and consensus search (Consensus) to map what’s already been done. Relevance asks who benefits and whether the topic connects to current priorities; science journalism (e.g., ScienceAlert) can help identify what’s “hot” outside academia and what may attract funding. Feasibility requires planning for a realistic timeline (often ~2 years for core progress), including data access, equipment, expertise, ethics approval, and the ability to adjust scope when experiments fail. Strong ideas are pressure-tested through conversations with multiple people before committing.

How can a student judge whether a PhD question is “novel” without aiming for world-changing breakthroughs?

Novelty is framed as “a little bit new,” not a revolution. The practical method is to read the literature and check whether the exact question—or close variants—already exist. That means testing larger and smaller versions of the question, drilling into very specific sub-questions, and asking whether anyone has already answered them. Semantic search can help map prior work quickly: enter a research question into a tool like Elicit, review what appears, then repeat with another system such as Consensus to confirm what’s been covered and what remains unresolved. Over time, repeated gap-checking builds a sense for what counts as genuinely new.

Why is relevance treated as a separate criterion from novelty?

A question can be new yet still fail to gain traction if it doesn’t matter to anyone. Relevance focuses on who benefits from answering the question—inside academia and beyond. The transcript suggests using science journalism as a reality check for what’s currently salient: searching outlets like ScienceAlert for a topic (e.g., bats) can reveal current public concerns and emerging research storylines, such as links between bats and pandemics. The broader point is that funding and attention tend to follow “hot” areas, so relevance can influence long-term support for the research direction.

What does “feasible” mean in practice when planning a PhD topic?

Feasibility means the project can produce meaningful progress within the student’s actual time horizon, often around 2–3 years for the core work. Because PhDs involve repeated failure and iteration, the transcript argues that only about 1.5–2 years of effort is likely to be “successful” if failure is accounted for. Feasibility also depends on logistics: whether the student can access or create the needed data, whether the equipment and expertise exist, whether ethics approval is required and achievable, and whether paperwork and administrative steps fit the schedule. Scope must be sized so it’s not a weekend-sized task and not a decade-long program.

How should a student choose the right scope so the project survives when things go wrong?

The transcript recommends actively tuning scope. Start narrow enough to be manageable, then check whether it’s too narrow—if failures would leave the student stuck with no path forward, it’s too tight. If the scope is too broad to finish in the available time, it’s too wide. The goal is a “sweet spot” that leaves options: when experiments fail, the student can readjust focus rather than hitting a dead end. This is treated as an expected part of research, not a rare exception.

What role does feedback from others play in selecting a PhD topic?

The best topics are described as “pressure-tested” by multiple people before the student commits. That includes talking with the supervisor, other PhD students, postdocs, and people in the intended research group. The transcript warns against relying on a single perspective—either the supervisor’s preferences or the student’s own idea without external checks. Multiple rounds of questions and critique help refine the topic into something with clearer novelty, relevance, and a realistic path to completion.

Review Questions

What specific steps can you take to verify that a research question is novel rather than just “new to you”?
How would you evaluate feasibility using access to data, equipment, ethics approval, and a 2–3 year timeline?
What signs suggest your topic is too narrow or too broad, and how would you adjust scope to preserve options during failure?

Key Points

1
Students should not let a supervisor’s current interests automatically determine the PhD topic; the student should lead the search for a suitable research question.
2
Novelty is about finding a real research gap through literature review and testing variations of a question, not about aiming for world-changing breakthroughs.
3
Semantic search tools like Elicit and Consensus can help map existing work and identify what remains unanswered.
4
Relevance matters because beneficiaries and current priorities influence attention and funding; science journalism (e.g., ScienceAlert) can help gauge what’s timely.
5
Feasibility requires planning for what can be accomplished in about 2–3 years, accounting for repeated failure and iteration.
6
Scope should be tuned to a “sweet spot” that allows pivoting when results fail, avoiding both dead-end narrowness and unfinishable breadth.
7
Strong topic ideas should be pressure-tested through conversations with supervisors, peers, postdocs, and members of the target research group.

Highlights

Supervisors may steer students toward work that matches their own research, which can produce topics that are either too narrow or too broad—so students should take ownership of the research question.

Novelty doesn’t have to be dramatic; it’s enough to find a small contribution that hasn’t been done, verified by repeated literature checks and question variations.

Relevance can be assessed by looking at what science journalists are highlighting now, since funding and institutional interest often follow “hot” themes.

Feasibility should be judged against a 2–3 year reality, including data access, equipment, expertise, ethics approvals, and the paperwork burden.

The best topics are pressure-tested by multiple people before becoming the thesis foundation, not built in isolation.

Topics

PhD Topic Selection
Research Novelty
Research Relevance
Research Feasibility
Literature Review

Mentioned

Andy Stapleton