There are a number of lessons I learnt as an undergraduate which I wish somebody had told me much earlier.

Although each of you learnt different lessons with different notions of which way is the right, please share your findings and wisdom so that an eager reader would contemplate over your advice and would not regret avoiding later in life.

  • 5
    I think this question is essentially polling for opinion and thereby not a good fit here, albeit the answers being probably useful for many aspiring students.
    – posdef
    Oct 14, 2013 at 9:59
  • 4
    However, as this seems to be such a useful opinion question, the rare use of Community Wiki seems appropriate here.
    – Ben Norris
    Oct 20, 2013 at 10:31

6 Answers 6


Research can be great, but I learned it can be so hard.

First, Google "should I do a phd" and then "should I do a PhD in ". There are tons of people who succeeded and failed in academia and writing about it, about what's it going to be in Academia, and so on. Two examples:

Beyond that, here's a few things I've learned:

  • Being a researcher is a career job, not so unlike being a lawyer, a manager.

    • After all, a paper has to change the scientific knowledge on the topic. You're going to teach respected scientists something they didn't know and something they thought was wrong.
    • I hear that in life sciences, the career aspect can extend to people sabotaging fellow researchers (that is, competitors).
  • Being more intelligent does not (necessarily) mean being a better researcher. If you're admitted to a decent program, you're more than smart enough for research. Managing your work (in all senses) can be more important, even though supervisors can help you with (part of) this in the beginning. But it's also about discipline and character:
    • Choosing a good question to work on (supervisors can help) - the question should matter (or you should make it matter, if you can).
    • Keeping an overview of what you're doing is important: otherwise, you might find an excellent solution to a problem which in the end is the wrong one.
    • Working enough on it (keeping focus among distractions of grad school can be hard).
    • How do you choose which way to approach a problem? Some approaches might not ever work, but others will only work if you keep at them long enough. So you need to be tenacious, and be able to change mind, depending on the moment. You need to delude yourself that you'll succeed until you do.
  • A good supervisor is extremely important. Nowadays it's extremely hard to start doing research "by yourself". On the one hand, there's too much written knowledge to navigate (literature). On the other hand, there are often unspoken assumptions which you won't learn from reviewers. It helps if he's available enough for you.
  • A supervisor must also fit you, character-wise. You mustn't be friends, but you need to function together, and neither you nor your supervisor study how to occupy your position.
  • Communication is extremely important. You're supposed to learn something and then tell your peers (and the world). And nowadays, if your peers don't get it, it's your problem (because your paper gets rejected). Hence, learning technical writing was very important for me.

Of course this is what I learned, which is influenced by my weak points, my experience and my field — I'm saying this because I'm sure somebody will disagree. I'm a programming language researcher, so I invent and design stuff (programming language). And evaluating design is hard. So much that many ideas are not right or wrong per se - you can make it right or wrong by determined work (see also this). But other disciplines probably experience similar phenomena in different form.

  • 4
    "neither you nor your supervisor study how to occupy your position". I don't understand what this means. Dec 25, 2013 at 20:08

I've seen a lot of people have a bad time until they realized this:

  • your supervisor/adviser is somebody you should work with, who can guide and direct your research, but after not so much time, you, not the supervisor, become the expert in your topic.

    Supervisor is the more experienced researcher, and his expertise should definitely be taken into account. But, after months of research into the topic, you start knowing more about your specific problem than anybody else, and should not expect the supervisor to be a fail-safe book of infallible answers.

Another very important one that I'm happy was told to me:

  • you can't do research as a side-activity. In order to do it well, and to be happy... um... in your life, you have to love research in order to do research.

    There might be jobs that are not your dream jobs, but are worth having because of other personal benefits (e.g. proximity to where you live, good salary, etc.) which you can do full time and have a happy and full life beside them, even if you're not working your dreams. Research is just too demanding and too exhausting to do if it is not the thing you want to do.

  • 1
    There is considerable potential danger in a novice' thinking that they will "become expert" in a few months. Certainly this is infeasible in contemporary mathematics. If it is possible even in a meaningful relative sense, I'd claim one needs a better advisor. And another danger is that if one has the expectation of "being expert in a few months", and has the sense to observe that this doesn't happen, one can (falsely!) judge oneself a failure. In mathematics, it would be extremely foolish to imagine that one becomes "expert" after some months... Feb 27, 2014 at 19:02
  • 1
    @paulgarrett Maybe my wording was clumsy, but what I was talking about was "becoming an expert" in a certain problem/sub-(sub-sub-)-field assigned to you for your project (thesis, internship, whatever). It might be different in Mathematics, but with my field, even my short few-months internship projects in Computer Vision, while my adviser was always the far superior expert in the field and great for general advice, after a short while, I would know more about the specific project-related problems and challenges, feasible approaches than him/her.
    – penelope
    Feb 28, 2014 at 9:21

Two key pieces of advice:

  • Learn to discern. You will encounter a lot of new information as a graduate student. Some of it will be absolutely illuminating; much of it will be of at most tangential benefit. Other papers and ideas won't be worth the paper they're printed on (or the bits of hard drive space they take up). One of your biggest challenges will be to figure out what's worth knowing, and what isn't. This will take skill, practice, and guidance.

  • Learn how to learn. As a researcher, the odds that you will be doing the same thing throughout your career are asymptotically small. Your PhD should not be just about picking up skills for solving a particular problem. Ultimately, it should be about learning how to become an expert in a field you haven't seen before.


Networking is crucial (it's not only a game of skill and hard work).

It works in two ways:

  • most opportunities (positions, workshops, fellowships) spread organically,
  • for everything you need to provide 2-3 recommendation letters; if you don't know professors who are well-known and respected in the place where you apply to, it may strongly affect your chances.
  1. Don't fly solo. Having initiative, coming up with your own ideas and avenues of research is great, but find someone to take responsibility for you and your success - especially when it concerns funding and advocating for you in the department. Being a ronin grad student involves semi-recurring desperate searches for funding.
  2. Don't get too distracted by side projects - they're sometimes quite useful, but it's easy to get bogged down in new shiny things, and a collection of unrelated musings does not a dissertation make. Similarly, don't be afraid of presenting the same thing twice - continually submitting unrelated, novel projects just leaves a bunch of half-done things gathering dust.
  • 7
    Responsibility, perhaps. Funding, if possible. Mentorship, certainly. But "ownership"? You've got to be kidding.
    – JeffE
    Oct 13, 2013 at 15:47
  • 1
    @JeffE "Taking ownership of a problem" is, I think, a pretty commonly used colloquialism at this point. And is much shorter to write than "Responsibility, mentorship, a role in advocating for you in the department, and viewing you as a member of their group". Do you also object when someone says "These are my lab members"?
    – Fomite
    Oct 14, 2013 at 2:15
  • 3
    There's a difference between taking ownership of a problem and taking ownership of a person. "These are my lab members" means "These are members of my lab", not "I own these people."
    – JeffE
    Oct 14, 2013 at 13:20
  • Maybe it differers, but for me side projects turned out to be the most beneficial thing. Oct 14, 2013 at 14:24
  • @PiotrMigdal My side projects too had pleasant outcomes, but I likely spent more time on them than I should have.
    – Fomite
    Oct 14, 2013 at 14:59

You must log in to answer this question.

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