17

A PhD is supposed to be an "original contribution to knowledge". In my own PhD (in an area of Earth Science) I have so far focused on developing a particular method of computer modeling that will hopefully have a range of applications in my field. In short, my question is: Is the development of an original/novel method a contribution to knowledge worthy of a PhD?

The reason for asking this is that my supervisor seems concerned that I am drifting away from the 'pure science questions' in my field - i.e I ought to be trying to address a fundamental research question with this new model, rather than spending most of my time developing and testing a new method.

My instinct says that it should be just as valid, and browsing some of the other PhD theses from the department suggests that a few others have gone down this route. I also see plenty of papers published that are more 'methods-focused' than 'fundamental-question-addressing'. Personally, I get a lot more satisfaction from pursuing the former kind of research, even though I accept I am not directly answering 'the big questions' as my supervisor puts it.

On the other hand, I don't predict an entirely healthy academic relationship down the line if our opinion on what constitutes an interesting PhD thesis differs so fundamentally, and I also consider the possibility that I could just be plain wrong in my interpretation of "original contribution to knowledge".

  • 1
    I've developed millions of lines of code and dozens of new methods, published 3 peer reviewed papers, one in a major journal and my advisor still isn't happy. – user22461 Oct 1 '14 at 22:25
10

Is the development of an original/novel method a contribution to knowledge worthy of a PhD?

This depends largely on what you mean with "method". Coming up with a theoretically sound way to tackle a specific class of research problems better than currently possible is likely a good contribution (with potential for very high impact). If "method" means a technical improvement of an existing approach or tool, the contribution may be too technical.

However, as you indicate yourself, this really isn't the important practical question in your case. Even if we all here agree that a new method is a solid research outcome for a PhD student, what good is this to you if your advisor disagrees? At the end of the day, your research needs to make both, you and your advisor happy. Doing something that your advisor does not think is PhD-worthy will lead to no ends of trouble, no matter what we random people on the internet have said.

  • These top two answers were good, but I chose this one as you rightly point out that it's an advisor/student issue if we don't see eye to eye on what a PhD thesis should be. – Jay T. Oct 8 '14 at 19:20
15

While developing a method has merit within science, that method needs to be useful to answer a question which is not currently answerable. I too have worked primarily on methods development in graduate school, and got stung by this in some of my early work. So, while it's good to focus on the method, always have the question in front of you: "what do we want to know, that we can't know yet, that my method will be able to tell us?"

  • 5
    Alternatively, the method needs to be considerably more efficient. Just because there's a solution available, if it takes significant amounts of time to calculate (even using a computer) and your new method makes the work comparatively trivial... I'm thinking reducing Big-O for CS algorithms and such for "important" functions. – Doc Oct 2 '14 at 0:05
  • Yes, definitely. – Danny W. Oct 2 '14 at 2:02
  • @Doc I'm curious as to how you were "stung" by this approach - did you feel it affected your chances of progressing into an academic career? Or did you find that you weren't in a strong position for the viva/defence from focusing on method over fundamentals? – Jay T. Oct 8 '14 at 19:24
  • 1
    This looks like it was more directed to me? Anyway, in general it's much harder to get published in the "best" journals with only a method than it is with a method + new science. I think this is fair too - if you can't find a good use for your method, why should someone else try? – Danny W. Oct 8 '14 at 21:41
  • @DannyW. Yes, sorry, mis-tagged it! Yes I should clarify, my research isn't just 'method', but more: "Here's a new method x, and now we can use it to predict y much better than before." – Jay T. Oct 9 '14 at 17:54
6

Yes. This was the basis of my successfully completed PhD.

Scientific methods must be able to solve something that can not yet be solved, or be (as was my case) an innovative improvement on an existing methodology that is more accurate, more accessible and if possible, more inexpensive. The latter requires considerable validation of the data produced.

It is critical that the method be able to be scrutinized scientifically for repeatability and reliability.

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.