Back to the beginning

One advisor’s perspective: Have a growth mindset

By Tom Anderson

Here, I distill my experience into a few points worth keeping in mind as you undertake your PhD.

  • Research is hard. We are trying to do things that haven’t been done before, and we aren’t always sure about the best path to take. Mistakes are inevitable, going down blind alleys is inevitable, and from time to time you may even come to wonder whether you are suited for a research career. This is normal. Everyone in grad school goes through both ups and downs. We can almost guarantee that the best work you will do in your life is not what you are doing right now – no matter how important it feels to do it well – but what you will do after you graduate. Making mistakes is part of learning.

  • Be impatient. The other people working in your area are also smart and hard-working, so you are always at a risk of being scooped. Furthermore, grad school is a means to an end, not an end in itself. Being able to get projects all the way to quality completion, as rapidly as possible, is part of the job.

  • Be a self-starter. Although your advisor can help diagnose problems, ultimately you’ll need to self-diagnose. If you have gotten yourself into a rut where you feel you aren’t making progress (or as much progress as you’d like), self-diagnosing is a good first step.

  • Set priorities. It is good to be full of ideas, but it is not the number of projects you start that matters, it is the number and quality of the ones you finish. A good discipline to develop is to keep a list of all the things you could be working on and rank them with the help of your advisor. Then, only do the most important one until it is done and move on to the next. (This advice is right for most students. Some students like to have a few projects going at once, so that when they are stuck on, or tired of, one of them, the student can be productive by working on a different project. Managing multiple projects requires that you are organized and careful about taking notes to ease context switching.) You may find yourself adding to the list faster than you can complete things. That’s actually good. You may find that other people picked up on something on your list and got to it first. That’s okay, too. Had you split your time, you probably wouldn’t have gotten anything done fast enough to matter.

  • Beware both overconfidence and underconfidence. It is easy to feel like you don’t belong in grad school, that you are an imposter, that everyone else has it all figured out. Don’t stress out about this. Instead, have a growth mindset: failing at something doesn’t mean you are bad at it, it just means you are in the process of learning it. Grad school will force you to learn a set of skills that undergraduate colleges usually don’t teach and that are difficult to learn: how to pick a topic, how to write a research paper, how to sell someone on your research agenda, how to give a talk, how to find the holes in other people’s work when they are doing their best to hide them. Your advisor knows that these are hard to learn. On the other hand, overconfidence can be fatal; if you think you know something and you are wrong about that, it will be hard for your advisor to help you at all.

  • Manage risk. In research, you are trying to do something that no one else in the world knows how to do. It is essential to manage risk. Identify up front everything that could derail a project. Work on the riskiest part of the project -- the one that you are most worried about. (The riskiest aspect is not necessarily the hardest or most time-consuming part.) Confronting risk early is known as fail fast. It enables you to change your plans early, before wasting too much effort. You don’t want to discover at the end of a project that it won’t work. Of course, the future is an undiscovered country. Sometimes there are unavoidable risks, or risks worth taking.

  • The value of postmortems. Whenever you finish a project, you should ask yourself what you or your team could have done better, more quickly, etc. Of course, some mistakes are inevitable; when you are building something for the first time, there is a lot you won’t know. What you want to avoid in the future are those things that, in retrospect, were avoidable mistakes, where you could have done better.

  • Dealing with feedback. It is common the first time you submit a paper or talk about a project outside of your research group to get negative feedback. Don’t let this throw you. You shouldn’t be too surprised that it is hard to convince people of your ideas. What you may not realize is how common it is for even the smartest researchers to give initially negative feedback to some idea and then realize only later (sometimes much later) how misguided that feedback was. Unfortunately, this is often because the ideas were presented poorly. If in the meantime you gave up because of the feedback, what a loss! A concrete example of this is that the paper inventing zero knowledge proofs was rejected the first five times it was submitted. Unfortunately, research requires having a bit of a thick skin. On the other hand, a critique may be on target, at least in part. It is best to be able to welcome feedback rather than take it as a threat or personal rebuke. Assuming you can handle it (not as easy as it sounds), it is much better to know about an objection than not to know or, worse, be ignored.

  • Try to figure out how to do your advisor’s job. You may think that research projects are most efficient when there is a division of labor. In economics this is called the theory of comparative advantage. Everyone should do what they are relatively better at doing. While efficient in the short term, this allocation of responsibility can lead to your not learning essential skills, making it less than optimal over the long term. If your advisor is particularly good at doing something, that’s probably a really good thing for you to learn via observation and imitation and practice. Even if you aren’t heading for a faculty job, technical leadership in industry often requires skills similar to those held by faculty.

  • Your advisor won’t always be right, and they may not realize it. You should almost always listen to your advisor, but you should also know that they will sometimes be stone cold wrong about something. Ultimately, you are responsible for the quality of work you put your name to. If that means ignoring or going beyond the limitations of your advisor, then you need to be ready to do that. We’re advisors, not omniscient or perfect managers.

  • Develop a growth mindset. Suppose you try to master something new and difficult to learn. Objectively, you won’t be good at it at first, especially compared to others who may have had a head start in learning the skill. You could at that point say, “I’m bad at this, I can’t do this, I don’t want to do this.” But the more rational thing to do – and the advice you would probably give yourself if you were being the teacher in this scenario – is to try to avoid judging yourself. Fear is the mind-killer. Learning anything worthwhile takes time, and where you start doesn’t need to define where you end up, unless you let it.