45

My PhD work is on modeling a 3D printing process using a commercial FEM software. But I am not confident with basics of finite element theory, non linear Finite element analysis to be accurate, which I learnt by self study. My work was mostly application of established models, so I didn’t have to do any numerical work in my research. It's my final year (starting to write my thesis) and I am not happy with my level of expertise. I am freaking out. What do I do? Sad part is I am in 6th year of PhD.

  • 32
    Are you saying that you don't know enough to write the thesis or that you feel like a fraud for writing it without a deeper understanding? Those are quite different problems. – Buffy Feb 3 at 13:08
  • 5
    The latter. I feel that even though I have published or done some novel work, I don't feel satisfied with my level of knowledge. I feel as a fraud when writing about it without a deeper understanding. – user103945 Feb 3 at 13:14
  • 131
    This sounds like every doctoral student ever. – Strawberry Feb 3 at 16:11
  • 37
    "I have published or done some novel work". Thank you for your contribution to science in what appears to me a field that will be very important in the coming years. Six years of your life is a lot, and I'm sure the scientific community (and those outside it) will join me in thanking you for commitment---not just in terms of time, but mental energy and sacrifices. Now go get the credit you deserve! Well done, and good luck! :) – Sam T Feb 3 at 20:32
  • Comments are not for extended discussion; this conversation has been moved to chat. – eykanal Feb 6 at 14:09

10 Answers 10

85

Nobody knows everything. In many fields it is impossible to know everything about that field. In mathematics it was possible for an individual to know everything up through the end of the 19th century, and a few people did. Henri Poincaré was a likely candidate. But after that it soon became impossible as there was just too much. Computer Science has probably passed out of the era where it was possible to know it all, but it was (possibly) somewhere around the end of the 20th century.

There is a psychological effect called Imposter Syndrome in which you don't believe that you are as good as others know you to be. It affects many people, including many academics.

Learning is a life-long activity. I suggest that in the short term you just get on with it and write the thesis, working as you normally do, which I'm sure includes study. In the longer term you can, as you would likely do anyway, work to fill in those bits that you think you are missing.

But don't delay your work now until you think you are ripe. That point may never come, even though you are perfectly capable at each point at which it matters.

The feeling may never go away. But if you know it for what it is, it needn't inhibit you or become debilitating.

  • 12
    + for imposter syndrome. And + for "nobody knows everything". No man or woman is an island. If you know where your weaknesses are, when work comes up that requires that expertise, you reach out to your colleague who is strong there. My stats suck but I know when stats are needed and I get a collaborator with strong stats. Look at the link for impostor syndrome. It has some ideas to help people afflicted. – Willk Feb 3 at 17:20
  • 5
    @Willk Obviously "nobody knows everything", but OP is talking about basics of the field in which he/she published already. It's quite likely there will be questions about FEM in his/her defence, where no colleague can help out. – M. Stern Feb 3 at 17:37
  • The whole point of PhD study is to confront the unknown. Just go for it! – O. Jones Feb 4 at 21:56
  • 3
    Actually at the end of my PhD I thought I knew less of the subject than at the start, when I thought I knew everything. – Bernhard Feb 5 at 6:57
  • 4
    @Bernhard I believe that is the Dunning Kruger effect – profMamba Feb 5 at 11:18
18

It's not optimal but quite natural that during your PhD you always push forward and only learn what you really need at that moment. I assume you are going to write an introduction and a theory chapter. Use this chance to catch up with what you feel you missed. Do some exercises, work on proofs of important theorems in your field. I guess few weeks of concentrated labour will make you feel much more confident regarding the basics. It will pay off when it's time to defend your thesis!

  • 3
    He actually may want/need to skimp on the theory/intro especially if he is weak on it. I have a contrarian advice which is to write up all the experimental work first. Once done, then probably momentum pulls him to add some theory/intro. But good chance it is workmanlike and not beautiful. – guest Feb 3 at 17:38
  • 4
    @guest I'm not recommending an order here. I agree that getting started writing is important. The unresolved questions might cause some stomach ache until OP sits down to face them, though. – M. Stern Feb 3 at 17:45
  • 4
    I think it is the opposite. He is using asking questions to avoid getting down to business. The experimental work needs to be written down no matter what. Having something tangible produced will settle him down. The guy is six years in. We are past the dreamy explorational aspects of grad school. Time to turn the screws. (Different point of view, but worth considering.) – guest Feb 3 at 17:51
  • 1
    @guest Maybe I was not clear enough: It's good to start with anything that will be easier to write down. I don't think we disagree on this. But be careful what you assume: We don't know it's experimental work, we don't know OPs gender, we don't know whether he/she has any problems with writing or motivation! – M. Stern Feb 3 at 18:04
17

Just write up the details of the applied work you did. Sure, you see a gap in your fundamental understanding, but so be it.

There are analytical chemists who do good applied work but don't understand all the hairy physics math behind their instruments also. Sure, ideally they would. And to some extent they should. But that's just not the case.

Don't make excuses. Write up what you did. Sit down and type. ;-)

  • 5
    Don't make excuses. Write up what you did. Sit down and type. this is the right, or at least best answer, along with perhaps staying away from talking about writing on the internet. – uhoh Feb 4 at 6:45
6

Some elements you do need for your thesis are:

  • To be able to identify that you have done something novel
  • Why it is novel, i.e. understand and articulate what has been done before
  • Ideally some notion as to why it works.
  • Where there is room for improvement

So, I suggest you focus on the main story you have to tell and make sure you can defend the arguments presented. That does not require an in-depth understanding of all the foregoing theory, but might well require articulation of the core principles.

If you have a chapter of background material, clearly you need to understand it. You do not need to understand all the related theory work. You might well now see that there are various areas to further study later if you stay in academia.

2

You say you're about to write your dissertation: do you really know how LaTeX works? Or the TeX that underlies it? Or the OS of your computer under that? Or do you just know how to use them? Same applies to your FEM program, or to any tool in any field: you don't need to know how the tool works, internally.or how to build one. You just need to know how to use it to get results.

  • 8
    In some sense this is right, but in a deeper sense it is not. With TeX, you can easily judge whether you "got the results" you wanted: you just read the output. But (assuming you don't know Arabic, but are a TeX expert) suppose you were given the task of typesetting 100 pages of Arabic text in TeX. Would you be so confident you had "done it right" as you would be for 100 pages of English and math? I doubt it. The same is true of using nonlinear FEM analysis, in my experience - and the worst situation is when something "20% different from expected" looks plausible, but is really total garbage. – alephzero Feb 3 at 17:38
1

A little humility at the start of what may be your first real research project may not be a bad thing. Crossing the bridge from hard problems in textbooks to a research problem can be a difficult journey. With textbook problems, it is at least reasonable to assume that the question makes sense and has an answer. With a research problem it may take a few months before you make sense of the problem, and you don't know whether there is an answer. "If you know what you're doing, how long it will take, or how much it will cost, then you're not doing research."

But you should try to get over feeling "unworthy" of a Ph.D. Presumably, some group of people feels you have the potential to write a decent thesis, or you wouldn't be a position to try. It would be a fine thing if you win a Nobel Prize for your work, but the objective for now should be less ambitious. You need to find a worthwhile unsolved problem, get to understand what is known about the general area that's relevant, and start figuring out how you can know what's true and what's not. The amazing thing is that it's not usually possible to know how important your discoveries may turn out to be.

1

But I am not confident with basics of finite element theory

You've got several options

1) Get direction from your current advisor as to how to fill in the gaps

2) Take a class on this topic

3) Get a co-advisor specifically to fill in these gaps

0

As long as you show you have a general understanding of Finite Element Method and build a story around your application of it, you should be in a strong position to defend your work.

To be more specific - you already mentioned that your work is application of established models. This is absolutely fine if you provide appropriate references. Also, as others have mentioned, you are not expected to know everything so don't be so hard on yourself.

0

Your adviser's job is to tell you if you are not ready to do this. That is the adviser's job -- we're not even talking about a "psychological support" role but a simple professional judgement about whether you are capable.

If your adviser cannot even do that, find to another adviser. A good adviser is fact one thing that you cannot do without.

0

Freaking out might be a healthy feeling at this stage. I am not sure if that helps, but I was in your position once and I got my PhD after three years. I remember telling a friend the same thing as you just said (except the technical details were different), and he told me that everything would be alright because you're feeling that way. I am sure the things you just mentioned are not simple concepts and reaching the level to make novel contributions is not an easy road, but you'll get there eventually.

protected by eykanal Feb 4 at 21:55

Thank you for your interest in this question. Because it has attracted low-quality or spam answers that had to be removed, posting an answer now requires 10 reputation on this site (the association bonus does not count).

Would you like to answer one of these unanswered questions instead?