I have just finished my bachelor's degree in computer engineering. I am doing an REU with the professor who will be my adviser this fall when I enter grad school. Currently my future Adviser has me working on research that involves image processing. I was hoping to head in other directions in embedded computing and security.

From our weekly research meetings it seems like my adviser will be assigning me research topics through out my grad school career. Is this normal that an adviser would insist on determining the research topic of the PhD students under them. I have not pushed back on this yet but I am not prepared to go in depth in image processing. I am not a math major or CS major and that field gives me little interest.


As far as funding goes I am currently funded by and REU(research experience for undergraduates) and am unsure of where this funding comes from(NSF, department, professor). Here in 2 weeks though I will be funded by a Distinguished Academic Fellowship(DAF) that my university graduate school awarded me.

In order to get my DAF I had to write a research proposal. I proposed research into microprocessor, GPU, FPGA unification. As I understood it from those within my department I am not required to carry out that research as the proposal is more a way for the DAF committee to see that I can write a coherent proposal for later on in my academic career.

Concerning the union of my professors research interest and mine: He focuses on embedded system design for single specific applications. It just happens that in the current path of research he is onto it involves cameras. This then introduces the field of image processing into the project.

It seems that all his grad students are doing peripheral work on UAVs that will be coalesced into a UAV platform of some sort.

  • 30
    Something just does not feel right on your question. When you decided to have him as an advisor, didn't you discuss on what area you will focus? What is his area of expertise? If he is into image processing, it was unreasonable for you to expect him to do embedded computing or security.
    – Alexandros
    Commented Aug 1, 2014 at 22:03
  • 15
    I'm definitely a bit concerned as well: generally you pick an advisor because your research interests align pretty well. In many ways, you should never choose a grad program for its program, but for the specific labs you would want to be part of (i.e., the mentors).
    – Namey
    Commented Aug 1, 2014 at 22:49
  • 4
    An adviser essentially hand-picking your research for you? Get out while you still can!
    – JeffE
    Commented Aug 2, 2014 at 19:43
  • 3
    Why did you pick this person as your advisor, if you did not want to do the research that the advisor is interested in? Commented Aug 3, 2014 at 1:29
  • 3
    @EnthusiasticStudent Because an advisor handpicking your research is like an instructor doing your homework for you. You might still get the degree, if you can stand the boredom, but you'll have missed the opportunity that the degree supposedly represents. In short: It's your degree; you have to hunt it down and kill it. The hunt is the point, not the degree.
    – JeffE
    Commented Aug 3, 2014 at 3:35

5 Answers 5


This depends very much on:

  1. How much interests and your advisor's interests align
  2. Where your funding comes from

First, interest alignment: Starting a grad program, you find an advisor with interests that overlap with your own, do the foundational things that need to get done, then expand on it with your own ideas. Later, when you are pitching grants, very much the same thing but with program officers: you have a great idea, and you find a way to dovetail it with their calls for proposals. Finding the common ground between your research and others' research is a fundamental academic skill.

Second, money: if you are funded through a grant that needs image processing research done, do you think that your advisor or the funding sponsor will give you a big thumbs up when you say, "Actually, I have some great ideas for embedded security devices I would rather work on." Nope. If you were in their shoes, shelling out $30-60k on a student for a year, wouldn't you expect them to help get the grant completed successfully? If you want full control over your research, become independently wealthy first.

Successfully doing your own research in academia, in grad school and beyond, involves a lot of meeting in the middle. Given the huge disparity of research topics you mentioned (image processing versus security), I genuinely wonder about the overlap of interests. If there really isn't much overlap, it might be worth looking for a different advisor. You won't become an expert in one field by being mentored by a guy in another field, even if your advisor lets you run wild with any idea that comes into your head.


Since more info was added about the funding and interest overlap, I can note a bit more here. It sounds like your funding is relatively portable and not tied to a project. In that case, you should have pretty strong control over what you want to do. You should not be using this control to just pitch ideas and run with them. The benefit of a PhD program (as opposed to learning how to research while living in a shack) is to apprentice with people who are experts in what you want to do.

However, having your own "batteries included" funding means you should be able to match with the best advisor for your interests. It sounds like you need to discuss the issue of overlapping interests with your current advisor. If you don't want to go in the same direction as the lab, maybe they can suggest another group who does something that is a better fit. Alternatively, maybe they will hit topics that you find more interesting during other phases of projects. It's not uncommon to switch advisors in the first year: I knew a professor in undergrad who actually recommended avoiding even starting in a given lab, so you could just meet with professors to find the best match.

  • The funding situation is unfortunate. Imho, finding a good problem and doing independent research is a hallmark feature of a PhD. "Here's a problem, solve it" can degenerate into a string of master theses. (Of course, if you want to work on problems completely orthogonal you your advisor's interests, you have a problem.)
    – Raphael
    Commented Aug 2, 2014 at 10:42
  • 3
    @Raphael: Do you really think so? In my field (mathematics), I think it's extremely rare for students to successfully find their own thesis problems. A major role of the advisor is to have a sufficient sense of the field to know about problems that are interesting and significant yet at an appropriate level for a student. Commented Aug 2, 2014 at 14:12
  • 2
    @NateEldredge I think it has to be a combination of both. The initiative should come from the "student" with the advisor controlling for scope. Imho, general interest ("significance" or "relevance") are not at all important for a PhD; it's about learning how to do (good) research, not the results themselves. (It's nice if both coincide, of course.)
    – Raphael
    Commented Aug 2, 2014 at 16:48
  • 2
    @NateEldredge There is a very broad spectrum between "Here's a problem, solve it." and "Find your own problem."
    – JeffE
    Commented Aug 2, 2014 at 19:45
  • 2
    @Raphael: for students in math who want to continue in academia and be competitive for postdocs and/or positions at top schools, the significance of their thesis research is quite important. Here "significance" is measured by the other researchers in their area not at their institution. Perhaps it is different in other fields. Commented Aug 3, 2014 at 2:11

In many fields, it is common for the advisor to "suggest" research projects to graduate students (including PhD students). In my own area of mathematics, for example, this is the case.

Of course, there is give and take between the student and the advisor about exactly what project the student will work on, and one can never know the outcome of a research project before it is done.

It is a well known phenomenon in math that some advisors have a vision for a long-term research program with many required intermediate results, each of which they will assign to a different PhD student as a thesis project. Depending on how many students they advise, not all the students may work on these results, but many of them will.

On the other hand, there are some advisors who give students more flexibility to choose a thesis project. But the project will always need to be in an area where the advisor has real expertise - otherwise, the advisor is not a good choice for the project.

It is also worth keeping in mind that the goal of graduate education is to get students to a point where, at the end, they are able to select their own research projects and carry them out independently. This does not mean that the majority of students are already able to do so when they enter graduate school. It is easy for a graduate student to take on more than they expect, if they choose a thesis project unwisely.

Ideally, you will find an advisor (1) who is very good at advising; (2) who you can work with successfully; (3) who works on topics you are very interested in; and (4) whose students have a history of good jobs after graduation. But sometimes it is not possible to achieve all of (1)-(4), in which case you have to compromise. For some students I have known, giving up on (3) was easy - after all, working on something long enough often makes it interesting. For others, not having (3) would be a deal breaker.

  • 2
    "This does not mean that the majority of students are already able to do so when they enter graduate school." +1 on this. Judiciously selecting topics is one of the most advanced academic skills.
    – Namey
    Commented Aug 4, 2014 at 4:06

I think the short answer should be "NO", but there are a lot of stuff which have to be taken in consideration.

Is it clear on what type of project you will contribute with your research, or is it a research that is not on the frames of any project?

If it is a research within the frames of a project which has got funding based on its documentation, you will have to work on what that research asks you to. And it is up to you to decide if you want to work on that topic or not.

In both cases, I think that the field of research and the expected outcome of your research should be defined even before the research starts. Of course along the path you might hit obstacles and look at different stuff, but not to the extend of completely switching fields.

Looking at your description it seems that you are working on completely off-topic stuff.

Bear in mind that the Adviser might be testing you, and see your reaction. Maybe if you don't react and accept to do everything that he asks you to, then he will continue with the same later on. Also, he might be testing your patience.

In any case, I strongly recommend that you have a very thorough and mature discussion with your adviser before starting to work on the research. And make sure that this discussion will result in "clearly" defined research goals, at least a well defined topic to focus on!

In general, you are supposed to know the fields of interest of your adviser and expect that he will ask you to do work on those fields.

image processing, security, and embedded are very broad definitions and (at least to me) look very unrelated especially when the granularity of a phd research is considered.

  • 1
    "[T]he expected outcome of your research should be defined even before the research starts": if you know what the expected outcome should be, it's not research!
    – aeismail
    Commented Aug 2, 2014 at 19:46
  • @aeismail I can not understand. Specially in MSc thesis, students/professors do not expect a new theorem to be reached. They work on one application of a theorem and try to solve a problem. They know what they want to solve and they almost know where they are going. For instance, if we are solving a differential equation, we know the problem and we know the solution should be a function, but the process to find that solution is unknown... I think research is not defined by where researcher is going. It is the process that researcher works on to connect a problem to its solution.
    – enthu
    Commented Aug 2, 2014 at 21:47
  • @aeismail maybe I've done a poor choice of words there, but by saying that sentence I did not allude to knowing beforehand the final result of the research, but rather considering the possible outcome(s). Commented Aug 2, 2014 at 22:39
  • 3
    @aeismail: won't that depend on the field? In math, if everyone expects a certain result to hold, but I am the first to prove it, that still counts as research. For example, everyone expects the "Riemann Hypothesis" to be true, but nobody can prove it, and if someone proves it they will be famous. Commented Aug 3, 2014 at 2:43
  • I am also confused by aesimail's comment. In most experimental fields, you make a hypothesis before you do the research, which is exactly that: proposing what you think will happen. Then you see if it did happen and report it. Often, the ones that don't match what you expected are the most interesting.
    – Namey
    Commented Aug 4, 2014 at 4:09

No, it isn't normal or desirable. The point of getting a PhD is to learn how to conduct research, and part of that is learning how to choose research topics. Your advisor should suggest topics and steer you away from bad ones and toward good ones, but summarily giving you research topics for the entirety of your time in grad school is doing you a disservice. Find a different advisor. (In case it helps, my advisor did the same thing: He presented me with a list of topics he found in a paper somewhere, told me to choose one from them, and refused to even entertain my suggestions when I asked for something else to work on. It was one of many reasons why I got nothing out of my time in grad school, and I don't want you to have to go through something similar.)


The answer is YES, sadly. If you do not like it, you got to talk to him/her, or get another advisor if no agreement can be reached.

  • This comment gives me a sadface. I would definitely not give an emphatic "YES" to an advisor assigning research topics ad infinitum. Students doing research should be gaining skills from it (and skills that they will find useful for their goals). A student shouldn't be forced to do research that doesn't match their path: if the mismatch is really bad, the student should mention it and the advisor should support them in finding a better match.
    – Namey
    Commented Aug 5, 2014 at 8:18
  • You're right. It is not a good idea to enforce slavery research. But it is common practice. Just talk to the advisor about your plans and weather he/she supports you.
    – Rene
    Commented Aug 7, 2014 at 10:31

You must log in to answer this question.

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