It’s getting to be that time of year. Grad school applications are due, acceptances and rejections start filtering back, and anxiety is in the air. I haven’t completed grad school yet, but I have chosen advisors a few times for various projects, some of which have worked out better than others.
Here is the abbreviated version:
Questions to ask yourself before choosing an advisor:
Do you think what the potential advisor is working on is interesting and important?
Can you easily understand each other when having technical discussions?
Does the potential advisor value the sort of work you want to do?
Are there opportunities to interact with and learn from many different people?
Is the school well-regarded, organized, and collegial?
And here is the long version:
The most important quality to have in an advisor is someone you can have technical conversations with. You should be able to understand each other easily. When you ask a question, your advisor should be able to answer the question in a way that indicates they understood what you were asking, and when your advisor asks you something you should be able to understand why they think it’s a concern. Communication works best if both advisor and student are good listeners, but more fundamentally, it’s about having similar approaches to problems and similar scientific values. It’s best if you have worked with someone before agreeing to be their advisee in order to figure out if you click scientifically. Some fields recognize this and require people to do lab rotations in different groups before choosing an advisor. Unfortunately, many times people have to choose their advisor without working for them.
All is not lost though- there are other ways to get clues as to how you may work with someone without actually working for them. First of all, the topics people work on are often good clues to their personalities. If you love producing great data, you’ll probably want to work for someone who values experimental work. If you’re interested in elegant models, you’ll want to work for someone who appreciates that sort of cleverness and mathematical rigor. Bad advisor/advisee fits happen when the advisee is, say, interested in producing an all-encompassing mathematical model to describe a system, but the advisor is more impressed by good data and running elegant experiments. People with different approaches to problems can be good colleagues, but the cases of this I’ve seen in advisor/advisee situations have not worked out very well for the advisee.
Secondly, never agree to work for someone without having a technical conversation with them. Ask prospective advisors questions you have from reading their papers (you have read their papers, right?), or where they see the lab going in the next five years, or talk about a project you’ve worked on. Make sure questions are asked and answered, discuss both details and big picture issues. Think about if you’re going to be able to learn from this person. I’ve had a few professors who, despite being clearly intelligent and well meaning, approach problems so differently from me it’s hard for me to learn from them. Don’t pick one of these people as your advisor! They can be great people to talk to about career advice or to get advice from if you’re stuck on a particular problem on their area of expertise, but you want your advisor to be someone that is easy for you to learn from.
Third, don’t pick a lab that is scientifically isolated. The situation you want to avoid is there being just you, your advisor, and maybe a few other grad students around. It’s not good for the science or your sanity to just work with the same few people all the time. Luckily, this is easy to avoid. You can choose a big lab, or a lab that has a demonstrated track record of sending students to conferences, or a school where many professors have an open door policy and there are other professors that do related work, or a lab that has more than one PI, or a lab that has a close collaboration with another lab, or a co-advising situation, etc.
Lastly, there is the school itself. In the best case, the school is well-regarded, organized, and collegial. Many schools are lacking in at least one of these respects. I think it’s important to assess the school on these merits, but being lacking is not necessarily a deal-breaker. It’s just a way to think about what will be important to pay attention to as you go through the grad school process. For example, if the school is not organized, you should step up your personal organization to make sure deadlines are met and requirements are filled. These qualities are mostly relevant if they keep students from graduating and finding jobs. If the school is so disorganized students never graduate in a timely fashion, so uncollegial that grad students are used as pawns in some larger agenda, or so poorly regarded that even the best grad students from the best labs can’t find jobs, those are signs to run despite a good fit with an individual potential advisor. The best way to find out this information is to talk to current grad students, who are often remarkably forthcoming about these deficiencies.
Too often the advice on choosing advisors that I see is all based on how famous the advisor is. It doesn't matter how famous the advisor is if you don't impress them because you're not interested or good at what they value most, or if you don't complete your degree because the department is disorganized and full of backstabbers. When picking advisors the ultimate question is: which one will get me out the door with the best career prospects?
