May I ask if it is a good practice for PhD students in STEM to have supervisors with no expertise in the students' field? I wonder to what extent should a PhD supervisor in the STEM knows about his/her students' research. In addition, I know some PhD supervisors with quite different fields (even in different disciplines) from their students' but they usually say that it is their students' responsibility to do research (as a completely independent researcher) without the supervisors' help in domain knowledge. I wonder if it is true. If so, why would PhD students need supervisors? Any ideas about this will be greatly appreciated.

In terms of some questions from the chatting zone, I think field in this case is a sub-class of a discipline. In other words, a supervisor may have different specific research interests from their students' but they are delved into the same field and the same discipline. For example, a supervisor and his/her student are both delved into econometrics, so they at least share the same paradigm when doing research in the econometrics despite that they may face different problems in specific contexts. Please correct me if there exist some problems/errors in my statement.

  • There is a question in the same context but I think my questions are definitely different, so please don't close this question. Please find academia.stackexchange.com/questions/52553/… – Eric94 Jun 3 at 17:30
  • 4
    In what situation would having an advisor with no experience in an advisee's area of exploration be beneficial? The student ought to end up deeper in the area than his or her advisor has been, but that's not what you're asking. Probably would be a poor outcome for a student hoping to do a thesis in numerical analysis to be studying under a number theorist. – A rural reader Jun 3 at 17:31
  • 6
    I think whenever this question comes up there is a big discrepancy in how different people think about the word "field". See also en.wikipedia.org/wiki/Lumpers_and_splitters What is a field to you? What is a discipline? – Bryan Krause Jun 3 at 17:33
  • 1
    Some researchers recruit from outside their fields, yes. Pretty common scenario for a research molecular biologist to recruit software developers, for example, to round out his or her interests and needs. But is it beneficial to the student? Well that's a different issue. – A rural reader Jun 3 at 17:38
  • 2
    @Aruralreader: Yes, that's an entirely different question. In my experience, what is not going to work if the needs are totally different: e.g. someone who wants to do a PhD in computer science needs to do cutting edge research in computer science. A molecular biology group may have application problems at hand that need such cutting edge computer science research, in which case it may be a good fit. OTOH, they may need quite the opposite: someone who implements methods that are known to work reliably. And that would be a misfit that is a recipe for disaster in terms of a PhD project. – cbeleites unhappy with SX Jun 3 at 19:37

Of course there are exceptions, but overall, the answer to your question is NO. Your supervisor has expertise in certain fields and will be more helpful if you work in that field.

I know some PhD supervisors with quite different fields (even in different disciplines) from their students' but they usually say that it is their students' responsibility to do research (as a completely independent researcher) without the supervisors' help in domain knowledge.

I consider that a huge red flag and would stay far, far away from a supervisor who says that. In my experience, supervisors who say that are just trying to avoid holding any kind of responsibility or accountability. I would like to know whether any of their students manage to find good academic jobs, or whether their students enjoyed their time in graduate school. There are graduate students who lament their PhD advisor decision and say, "I was completely on my own." That is not a good thing!

Being independent doesn't mean that you can't get access to a supervisor's knowledge, or that you can somehow do research without access to domain knowledge. In fact, researchers are constantly asking other researchers questions and getting help. In the acknowledgments section of a paper, researchers will say something like, "I would like to thank Person X for answering questions I had about topic Y." Yet the authors still did the majority of the thinking and work. They may have come up with the questions they addressed in their paper, and they may be exploring areas that have been neglected. They are getting help without anyone holding their hand through it - I think of that as being an independent researcher. My background is in pure math. To prove a theorem, you have to know a lot of theorems. Not all the theorems you need are written down. There will be a lot of hindrances if you can't talk to people and ask. Later on your advisor will write you recommendation letters. How will they write a good letter if they know nothing about your work? You are shooting yourself in the foot if you work with an advisor with no expertise in your subject.

  • 2
    "You are shooting yourself in the foot if you work with an advisor with no expertise in your subject." To add to this anecdotally, as a PhD student I can count dozens of times where my advisor's expertise has saved me months of unnecessary work. There were many dead-ends that I avoided because my advisor had the intuition and foresight to point out potential roadblocks with my ideas, and this kind of foresight only comes with years of experience in the field – Superbee Jun 4 at 17:08


If so, why would PhD students need supervisors?

is a key to the situation.

From what I've seen, there are substantial differences on how this question would be answered in different (academic) cultures.

E.g. I was told/reminded by my professor early on during my PhD research* that I was a fully qualified chemist - with all professional rights and duties that chemists have by my Diplom (= MSc; here BSc and MSc are usually prerequisites for a PhD). The formal description of Diplom thesis is a 6 months research project that the candidate performs on their own (after "guided" research in the form of obligatory research internships).
Obviously, with a fresh Diplom, one isn't an experienced scientist. But I was largely free to decide my approaches and the discussions we had were typically the group-public discussions similar to the discussions with more experienced researchers (postdocs, PIs) in the group. I guess my manuscripts did need more rounds of polishing the text, though.
From that point of view, it is perfectly fine to have a PhD "student" coming from another field because the group/larger project needs their professional expertise. And it is also fine to expect the student to be able to get all information in their field they need (including contacting people).

There are other academic traditions that see the PhD as step of the professional qualification. And in particular where you can enter a PhD program with a fresh BSc, it doesn't make sense to expect the level of professional independence I described above.

I've been involved with rather interdisciplinary research, where different people are needed to bring several professional backgrounds (fields) together to tackle the project. It would be rather unlikely to find a professor who is expert in all those fields.

That being said, it does not make sense if there is no relation whatsoever between the PhD researcher's profession/field and the thesis on the one hand and the professor's field/expertise and the PhD project on the other hand.
But the overlap between the professor's background and the student's background may not be that large. The professor can in any way supervise the student on the general aspects of research, and can take care of their part to the project/paper.

In a more Phd-is-still-studying culture, the student may have co-supervisors that cover the (sub)fields not covered by the expertise available in their group.

It's certainly a difficult situation if you are in a lone-wolf position with your profession during your PhD. However, there are some professions where this is the typical working situation (e.g. for me as chemometrician this is certainly the case) and one may argue that the sooner one learns how to deal with this productively the better.

(This is one of the few aspects of professional life where I think the move to online seminar formats due to the COVID situation has lasting potential for improvements - while an online seminar IMHO does not offer the same networking possibilities as an offline seminar, such a student can now at least stay somewhat connected also to their profession, whereas before the decision which conference to attend may have always had the answer "application conference" before.)

BTW, I did my PhD thesis in my orginal profession (i.e. chemistry) - but the specialization into statistics/data analysis chemometrics got me to the level of what was around in the group during my Diplom thesis and when I started my PhD my professor I was not looking for guidance (as opposed to critical questions of the approach I'd chosen) in that aspect in my group any more.

So from that experience, I'd say that you can easily end up in the situation described in the question regardless of matching fields.

What is in my experience also a recipe for disaster: if the management/supervision style doesn't match the level of independence. I.e., as I said it is IMHO fine to recruit a PhD researcher because they bring certain professional expertise that is not available in the group. But that needs to be clearly communicated so the PhD researcher knows what is expected of them and it needs to be accompanied by treating them on eye level as fully qualified professional - it doesn't mix with a supervision style that would be adequate for guided research done by a student (which could anyways not be provided without domain-specific knowledge of the supervisor).

*in my native language, the term for "Phd Student" = Doktorand doesn't have a connotation of being a student, and indeed you are not required to enroll as student when doing a PhD.


On top of agreeing with the other answers, I'd like to offer a bit different perspective. You might want to narrow down your specialization, because STEM as you write is very broad. In physics, most of the PhD positions are funded from research grants, and academic staff members to whom these grants are awarded become supervisors of the phd students hired within those grants. Obviously, since you have to write an strong application for a grant, you are likely to plan research in your field of expertise, and then the "field of expertise" becomes the same for the supervisor and the phd student.


In mathematics, in the U.S., I'd think that it would be irresponsible of an advisor to attempt to supervise a thesis project completely outside their scholarship, unless the student was very, very unusual.

Keyword is "irresponsible", I think, though, yes, conceivably unethical...

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.