7

Ok here is the situation: My advisor gave me a paper about new model X. After reading and researching in X, I became very familiar with X and had many interesting ideas to extend its capabilities.

One thing I noticed with my advisor is the lack of deep analysis over my ideas. By this, I mean a roadmap on how to get these ideas published, what are the things I need to polish/sharp/add to get it (most likely) accepted in top conference/journal.

I handed him one draft of a paper, he didn't even doubt its findings; all his concern is to get it published somewhere in ranked C conferences; all he did was praising my writing (which is poor IMHO). What makes it even worse is the fact that I have seen some papers published in top conferences/journals with the same ideas I have (actually three ideas).

This makes me wonder: Does he really know X?

if yes, then why he did not provide constructive ways to sharp the ideas to get it published in a top conference? or maybe I am asking for something that he can't offer!

  • 9
    Does it matter? (Part of) The point of a PhD is for you to become an expert on your question. You should expect to know more about it than your supervisor, particularly if you devised the project from a paper you read (as you seem to imply, which is a great thing for you to be doing). – Jessica B Oct 18 '14 at 8:30
  • 3
    @JessicaB There is a 2 years learning curve here. I do not think I can publish in top venues on my own (I do have my own limitations which I think its the advisor's job to help me strengthen my skills). – seteropere Oct 18 '14 at 8:59
  • 4
    Keep in mind that people can become faculty without any training in how to be a good adviser. For this reason, I recommend having multiple mentors with different strengths. – Anonymous Physicist Oct 18 '14 at 18:10
  • @JessicaB Yes, it matters. If your supervisor doesn't have any idea about the work you are doing then all sorts disagreement will begin to pop up. – user May 11 '15 at 1:50
11

I suppose it is rather hard to give a satisfying answer to this question with the amount of information available. I mean, it's hard to estimate all the unknowns in this particular case.

if yes, then why he did not provide constructive ways to sharp the ideas to get it published in a top conference?

One could speculate in so many different ways; that he does not have the necessary grasp of the field to give meaningful critique is one speculation. Here's another plausible explanation; maybe he wants you to try flying on your own a bit, as a sink-or-swim approach. My supervisor did (and still does) this very often, I have almost too much freedom as to how I write, and what I write and to where I submit, then when the replies and reviews come in we sit down and improve on what I have done. Could be a similar approach

But to come back the question of how to know whether the supervisor is knowledgeable in a particular subject, here are some things that could be useful in the general case:

  • checking your supervisors publication record. Maybe the s/he has published articles on about this model before you even came to the group. Or maybe he was involved with an older but related model Y?

  • inquire you supervisor for his interest in the subject. If s/he gives you a paper about model X, all of a sudden then there is certainly a goal with that. To-the-point, effective communication might help you figure out his/her grasp of the field.

EDIT: In reply to the comment:

if I have too much freedom as to how I write, and what I write and to where I submit, then whats the point of supervision!

My supervisor often says that his job is to teach me how to do research, not provide me with knowledge or make sure I become successful/famous/rich... Your supervisor has ZERO responsibility in providing you the means to publish in top venues, IMHO. He merely has a responsibility to make sure you get the tools you need to be able to do the research you want to carry out, later on in life.

As a side comment, I think your perspective to research might cause your frustration with your supervisor. Getting published in top venues is not a birth-right to all hard working grad students; and quite frankly, as a grad student where your articles get published have little to do with you or your work (at least in biomedical research).

  • He had never published on the area. He never worked on something closely related. "I have almost too much freedom as to how I write, and what I write and to where I submit" then whats the point of supervision! – seteropere Oct 18 '14 at 9:02
  • 6
    I said part of the point of a PhD is for you to become an expert in your field. Another point is for you to learn how to be a researcher (which includes the process of creating and answering questions yourself). While it is generally helpful to learn that from someone who knows well how your field works, it is not entirely necessary. It's not unreasonable for your supervisor not to know about one of your research interests (or to have misjudged where you would end up). If you think your supervisor can't help you enough, perhaps ask someone else for guidance eg the author of the original paper? – Jessica B Oct 18 '14 at 10:39
  • 7
    @seteropere if you wanted an expert in some area to be your supervisor, then why did you pick someone who has never published on the area or even something closely related? That being said, if you want to publish something on topic X, then it's entirely expected and normal that you should know more about X than your supervisor and not the other way around; and the input of the supervisor would be about the general academic process, style, contents; about research process and the wider field, but not on the actual topic of your paper. – Peteris Oct 18 '14 at 11:00
  • @Peteris Answering your question: I knew nothing about the area before starting my PhD. For your comment, when I read a theoretic IJCAI or AAAI paper, I can't imagine its the student who did it, at least not 100% – seteropere Oct 18 '14 at 19:41
  • 4
    @seteropere if you're not the #1 most informed person in the world about the narrow particular topic of your paper, then either you're not ready to write the paper, or the topic of the paper isn't scientifically novel and thus not worthy of publishing in most venues. If you're co-authoring a paper where the main contribution is from your supervisor, then it's resonable for him/her to drive the scientific agenda; but otherwise the point of PhD (and an absolute requirement for graduation) is to get to a stage where you can be the main author of papers and inform others, not vice versa. – Peteris Oct 19 '14 at 6:50
4

This is more an extended comment, but here goes.

My PhD advisor had me work on a topic (in applied probability) that he knew little about. I survived and got a PhD, but it was not pleasant. He's seems to have thought that the area was much more interesting/useful/productive than it was, and had wildly unrealistic/inflated expectations about applying it to mathematical finance, which unsurprisingly came to nothing in the end. I suppose in part because he didn't know much about it. It is a pretty obscure branch of applied probability that enjoyed a brief wave of popularity in the 90s, and quickly went cold again because, as it turns out, the area has formidable and (thus far) largely unsurmountable technical difficulties. At any time this area has had a handful of people working on it worldwide. I think with some breakthroughs it could become a useful area, but nothing like this has happened so far.

I learned later from other people (I was pretty naive about how academia worked) that faculty may try to use their students to try to learn about new areas. I've no idea how common a practice this is, but I have no reason to believe it is uncommon. I recall that the people I talked to seemed to think it was quite reasonable. I think it is wildly irresponsible.

In any case, I suggest you don't let your advisor do your thinking for you. If you like the field you are working in, and think you can be productive in it, then by all means continue. However, if you think your advisor doesn't know what is going on, it is quite possible he doesn't. Researchers are only flesh and blood, they are not omnipotent. If he doesn't know what is going on, he will probably be of little help to you. It is possible for your advisor to learn about the topic alongside you, but you are the best judge of whether he is doing that. If he doesn't seem to be, he probably isn't. And frankly, it is not a good thing for a graduate student to have an advisor who views his PhD as a learning experience. Ideally a PhD topic should be well within the advisors area of expertise. If he does not seem inclined to go into details and talks about generalities, that is a bad sign, in my experience.

Bear in mind that finding a auxiliary advisor who knows more about the topic is a possibility, but if want to go in that direction, I'd do something about it sooner rather than later.

Other people have commented here that it is reasonable for you to end up knowing more about the topic than your advisor does. That is a fair point, but that should not be the case at the outset. Maybe that is Ok after a few years of you digging into the topic. And really, it depends on you. If you are comfortable going it mostly alone in the topic, perhaps communicating occasionally with other researchers, then that is Ok. If you are not, then it could end up being a bit of a baptism of fire. We don't know what your topic is, exactly, so we can't say anything more about that. And research topics vary wildly in level of difficulty, of course.

  • 4
    "faculty may try to use their students to try to learn about new areas" I think it is absolutely essential for a lab to reach out to new subfields, topics and techniques, and the only way this realistically happens is by getting PhD students and postdocs that work on this. Why do you consider this problematic? – xLeitix Oct 18 '14 at 16:53
  • 4
    @xLeitix: I should think that is obvious. If the researcher is not knowledgeable in the area, he may not be competent to guide the student. This depends on cases, of course. As they say, the devil is in the details. As I described in my answer, my own experience is that this can go badly wrong. This approach with postdocs is less problematic than students. They have more experience, for one thing. Also, you use the term "subfields". We aren't necessarily talking about "subfields", but possibly entirely different fields. – Faheem Mitha Oct 18 '14 at 17:00

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.