My question applies to Science, Technology, Engineering, and Math disciplines (STEM).

Q: By what process does a graduate student determine his or her research direction?

(Assume that the graduate student is in the first year and has no previous publications with this supervisor.)

I can imagine several scenarios:

  • The graduate student proposes a research topic and then their supervisor evaluates it for originality and feasibility. In this scenario, the graduate student has the foresight to know what research directions are promising and appropriate.

  • The graduate student performs an extensive literature review. In this process, the student finds a challenging and/or interesting problem. (I am not sure how much effort this requires. I also don't know how to identify a problem as "interesting" or "challenging".)

  • The supervisor comes up with an interesting research problem or question and then presents it to the graduate student. (This appears to be a common pattern, given the plethora of complaints that I see on-line about uninteresting research.)

I would especially like to see responses from professors who are supervisors and also from graduate students.

  • 4
    Why the downvote?
    – xLeitix
    Commented Jul 15, 2015 at 7:35

5 Answers 5


Any of the scenarios you describe can happen, in practice. (I am in the United States; this may differ in other countries.) It depends on the student's interests and abilities, the advisor's interests and advising style, how broad advisor's interests are, whether the advisor has funding for a particular project and needs a student to work on that problem, and many other factors.

All three have happened within my own research group. (Another possibility that happened in my research group that you haven't mentioned: student initiates a collaboration with another researcher, who proposes a topic, and involves the advisor in the project as well.)

Actually, the scenarios you describe are not discrete scenarios, but a spectrum from "most student-driven" to "most advisor-driven." Some points on the spectrum include:

  • Student enters grad school with a topic in mind that's judged acceptable by advisor, and proceeds with that topic.
  • Student enters grad school with a topic in mind, that evolves and changes following advice from advisor.
  • Student enters grad school with a general idea and, together with guidance from advisor and extensive reading of literature, shapes it into a topic.
  • Student enters grad school without idea, advisor suggests general direction, and upon extensive reading and with guidance from advisor, student shapes it into a topic.
  • Advisor hands student idea, and student proceeds to execute idea. Of course, this is a PhD, so as student works on the topic, the student's mastery of the specific research will most probably come to exceed the advisor's, even though advisor originally proposed the idea.

Other scenarios include: student enters grad school with idea in mind and then changes directions completely in middle, student enters grad school with idea and changes advisors in middle but keeps idea, and many more..

  • 22
    Another one: "Student enters grad school level without a general idea and starts publishing based on opportunities - whatever comes up during work for a funded project, collaborating with colleagues, etc. After two or three years, student and supervisor check for common aspects of the publications so far and thereby define a topic that connects what was done so far and allows for more research along the same lines." Commented Jul 15, 2015 at 7:34
  • 3
    @O.R.Mapper This is a good one. Certainly how I did my PhD.
    – xLeitix
    Commented Jul 15, 2015 at 7:34

In addition to your proposals and ff524's answer, another common pattern is

Grad student is employed by a grant / funded project and needs to deliver more or less what the grant promised

At least in Europe, that actually seems to be the predominant mode of topic selection. Most students are externally funded by relatively concrete projects, and not all grants allow students to deviate far from the original plan (even if the plan sometimes is a bit outdated / boring by the time the student works on it, much to the chagrin of both, student and supervisor).

However, as O.R.Mapper correctly states:

In larger departments, this sometimes means there are several openings based on different projects. Topic selection then takes place by choosing the project that is most aligned with the future grad student's interests, or that can most likely be developed into a direction that seems interesting to the prospective grad student.

(that being said, it's not usual that the decision which project to select is entirely up to the grad student)

  • 8
    "not all grants allow students to deviate far from the original plan" - whether the grant allows it in writing, and whether the supervisor tolerates it (or even actively suggests it) can be entirely different. Commented Jul 15, 2015 at 7:41
  • @O.R.Mapper I think no grant allows you in writing to do something else with the money. However, some are in practice more lenient than others.
    – xLeitix
    Commented Jul 15, 2015 at 9:54
  • 4
    Well, some grants/grant applications are more explicit about what the funded project is about than others. Sometimes, one can manage to describe certain aspects vaguely enough to have a lot of leeway even when sticking exactly to what is written in the grant. That is where, to put it exaggeratedly, the bureaucratic dreamworld that requires a precise description of all research results that will be obtained during the 3 years to come, and the scientific reality that each result leads to new, unforeseen ideas that can completely alter the next research steps, meet even in official writing. Commented Jul 15, 2015 at 10:04
  • 1
    One aspect of this just came to my mind that you might want to explicitly mention: In larger departments, this sometimes means there are several openings based on different projects. Topic selection then takes place by choosing the project that is most aligned with the future grad student's interests, or that can most likely be developed into a direction that seems interesting to the prospective grad student. Commented Jul 15, 2015 at 11:51

I believe the best way of finding a topic is if the grad student does not have a fixed idea of the research project, but of his interests and abilities. So when asked about a topic I counter with obvious questions like "What are you interested in? What do you already know? Are there side conditions like funding, limited time?", but also with less obvious ones like "Do you want to work with a computer? Do you prefer doing stuff on your own or learning known things? Abstract or concrete?"

This requires some introspection on the side of the grad student, but thinking about your abilities is sometimes worthwhile. I personally believe that a large part of your perceived improvement over time comes not from getting better, but from learning what you are good at: You become more successful, because you learned to avoid problems that don't fit your working style.


In addition to what has already been said by ff524 and xLeitix, I think it also depends on what you mean by "research direction."

If you're talking about papers and projects to be completed during graduate school, including the dissertation - then yes, any of those processes work. I think it's most commonly a combination for them - for example, earlier in their graduate work students may do more things on suggestion from their advisor, but later in their graduate work they get increasingly creative and have a greater knowledge of the literature, and those are given more free rein to decide projects. Or it may be that certain projects and papers are ideas handed to the student from the PI or from other members of the lab, but other projects are independent ideas from the graduate student themselves.

It also depends on the PI themselves - some are more controlling than others, and many have varied ideas about what the role of a doctoral student is in a lab.

If you are more talking about a general research direction for your career, though, that's an agenda that really should be set by you - the doctoral student. Your advisor is there to guide you, but doing an extensive literature review and developing the foresight to know the gaps and hot areas in the field is what PhD work is all about.

For what it's worth, at the beginning of graduate school I wondered how professors were able to generate enough ideas to keep papers and grants flowing. By the end of graduate school, I myself was overflowing with ideas - I had too many to write papers on, and had to begin a spreadsheet to keep track of them. As time goes on you will learn more about the field, and you will also develop a network of collaborators.


Like any relationship, it really depends on the relationship that you have between you and your supervisor. Some will leave you completely alone to decide, others will guide you step-by-step. And it's a continuous spectrum!

You must log in to answer this question.

Not the answer you're looking for? Browse other questions tagged .